OPINION

Defendant Charles Lee Gillenwater, II, was charged with two counts of transmitting threatening interstate communications and one count of transmitting threatening communications by United States mail. After determining that Gillenwater was not competent to stand trial, the district court authorized the government to medicate Gillenwater involuntarily to render him competent to face the charges against him. While recognizing the important interests at stake for both the government and Gillenwater, we conclude that the district court did not err in authorizing Gil-lenwater’s involuntary medication.

I

Gillenwater once worked on the renovation of the Flamingo Hotel in Las Vegas, Nevada. Gillenwater believes that he and thousands of others were exposed to asbestos during that renovation. He also believes that the government allowed the exposure to occur and helped the hotel cover it up. And Gillenwater believes that government and hotel officials came after him when he tried to reveal the exposure and cover up.

In November 2011, Gillenwater was charged with two counts of transmitting threatening interstate communications, in violation of 18 U.S.C. § 875 (c), and one count of transmitting threatening communications by United States mail, in violation of 18 U.S.C. § 876 (c).

In the proceedings that followed, the government claimed that Gillenwater made graphic and disturbing threats against a number of government officials and employees and showed a possible intent and ability to carry them out. According to the government, Gillenwater came to the attention of federal authorities in July 2010, when an Occupational Safety and Health Administration (OSHA) employee reported receiving a threatening email from Gillenwater. Among other things, the email stated: “Violence is my primary means of communication and it usually takes the form of me choking somebody while screaming ‘CAN YOU HEAR ME NOW[?]’ ” ER 217-18.

Federal agents went to Gillenwater’s house to speak with him about the email. He met them with a gun in hand. But Gillenwater ultimately put the gun away and spoke with the agents. The agents warned him to refrain from sending threatening emails as it was illegal and could result in his arrest and prosecution. According to the government, Gillenwater acknowledged his understanding.

The government claimed that Gillenwa-ter was back at it two days later. He allegedly sent an email to Department of Labor (DOL) employees, which, among other things, advised them to “[l]ive in fear” and asked them “[d]o you really want to be between me and my enemy?” Id. at 218 . From there, the situation continued to escalate. According to the government, the volume and content of Gillenwater’s emails eventually prompted DOL to block his email address. And Gillenwater allegedly did not limit himself to sending threatening emails to OSHA and DOL employees. The government claimed that he also sent threatening emails to Senators and federal agents. The emails said things like:

Gillenwater admitted to sending at least some of the threatening emails, including one that threatened the life of an OSHA employee. But Gillenwater explained that he sent the emails to bring attention to the government’s misconduct in the Flamingo Hotel asbestos exposure and cover up.

Federal authorities arrested Gillenwa-ter. They found him with a gun and spare ammunition. According to the govern ment, the subsequent federal investigation revealed that Gillenwater had military training in the use of guns and that not all of his guns were accounted for at the time of his arrest. While in custody, Gillenwa-ter allegedly sent a threatening postcard to an OSHA employee.

The district court ultimately ordered a competency evaluation. Dr. Cynthia A. Low performed the evaluation and diagnosed Gillenwater with delusional disorder, persecutory type. As described by Dr. Low, delusional disorder is characterized by “the presence of one or more nonbiz-zare delusions that persist for at least a month.” Id. at 33 . And patients that suffer from the persecutory type believe that they are “being conspired against, cheated, spied on, followed, harassed, or obstructed in the pursuit of long-term goals.” Id. at 34 . Dr. Low concluded that the disorder could substantially impair Gil-lenwater’s ability to assist his attorney with his defense.

The district court found that Gillenwa-ter was not competent to stand trial and remanded him to federal custody to determine whether he could attain competency. Dr. Robert G. Lucking and Dr. Angela Walden Weaver performed a second competency evaluation of Gillenwater. They reached the same delusional disorder diagnosis as Dr. Low and concluded that Gillenwater was not competent to stand trial. Dr. Lucking and Dr. Weaver also concluded that Gillenwater could attain competency with medication, namely the antipsychotic drug haloperidol decanoate. Gillenwater refused medication.

The government moved to order involuntary medication pursuant to Sell v. United States, 539 U.S. 166 , 123 S.Ct. 2174 , 156 L.Ed.2d 197 (2003). The district court held three evidentiary hearings on the government’s motion. Consistent with his competency evaluation, Dr. Lucking testified that involuntary medication with haloperidol decanoate was substantially likely to render Gillenwater competent to stand trial. He also testified that involuntary medication with haloperidol decanoate was medically appropriate and unlikely to produce severe side effects. He rejected other treatment options, including voluntary psychotherapy.

Gillenwater then called Dr. C. Robert Cloninger to testify. Dr. Cloninger had not met with or examined Gillenwater. But he had reviewed the competency evaluations performed by Drs. Low, Lucking, and Weaver and related court filings. Based on those materials, Dr. Cloninger concluded that Gillenwater did in fact suffer from delusional disorder, as well as from a personality disorder with depressive and narcissistic features. But Dr. Cloninger disagreed with Dr. Lucking’s involuntary medication recommendation. He testified that, in his view, involuntary medication is not effective in treating delusional disorder and may even worsen the condition. Dr. Cloninger also testified that the off-label use of haloperidol decanoate is inappropriate if the medication is administered involuntarily and that halo-peridol decanoate may produce severe side effects. He instead recommended treating Gillenwater with voluntary psychotherapy.

On November 19, 2012, Chief Judge Peterson issued a lengthy order authorizing involuntary medication with haloperidol decanoate.

Gillenwater appealed both the November 19, 2012 involuntary medication order and the underlying competency determination. While his appeal of the November 19, 2012 involuntary medication order was pending, this court vacated the underlying competency determination, holding that Gillenwater was denied a sufficient opportunity to testify at his competency hearing. United States v. Gillenwater, 717 F.3d 1070, 1085 (9th Cir.2013). We remanded to the district court for a new competency-hearing.

On remand, the case was reassigned to Judge Suko. The parties agreed that a new competency evaluation was not necessary and that the district court could consider the evidence offered in the initial competency proceedings. The district court held an additional evidentiary hearing at which Gillenwater testified at length. On September 24, 2013, the district court issued an order determining that Gillenwater was not competent to stand trial and authorizing involuntary medication with haloperi-dol decanoate. In doing so, the district court incorporated the November 19, 2012 involuntary medication order.

Gillenwater appealed the September 24, 2013 involuntary medication order. We consolidated that appeal with his already pending appeal of the November 19, 2012 involuntary medication order. We have jurisdiction under the collateral order exception to 28 U.S.C. § 1291 . Sell, 539 U.S. at 176-77 , 123 S.Ct. 2174 .

II

A defendant “has a ‘significant’ constitutionally protected ‘liberty interest’ in ‘avoiding the unwanted administration of antipsychotic drugs.’ ” Id. at 178 , 123 S.Ct. 2174 (quoting Washington v. Harper, 494 U.S. 210, 221 , 110 S.Ct. 1028 , 108 L.Ed.2d 178 (1990)). For as the Supreme Court has explained, “[t]he forcible injection of medication into a nonconsenting person’s body represents a substantial interference with that person’s liberty.” Harper, 494 U.S. at 229 , 110 S.Ct. 1028 . Accordingly, the government may “medicate a defendant involuntarily for the purpose of rendering him competent to stand trial only in rare circumstances.” United States v. Ruiz-Gaxiola, 623 F.3d 684, 687 (9th Cir.2010).

When the government seeks to medicate a defendant involuntarily for competency purposes, it must establish by clear and convincing evidence the four Sell factors. See id. at 692. Those factors are: (1) “that important governmental interests are at stake”; (2) “that involuntary medication will significantly further ” those interests; (3) “that involuntary medication is necessary to further those interests”; and (4) “that administration of the drugs is medically appropriate.” Sell, 539 U.S. at 180-81 , 123 S.Ct. 2174 .

Gillenwater contends that the government did not meet its burden on all four Sell factors and that the district court’s order authorizing his involuntary medication must therefore be reversed. We review a district court’s determinations with respect to the first Sell factor de novo. See United States v. Hernandez-Vasquez, 513 F.3d 908, 915-16 (9th Cir.2008). And we review a district court’s determinations with respect to the remaining three Sell factors for clear error. See id.

A

We start with the first Sell factor — the important governmental interests factor. Needless to say, the government has an important interest “in bringing to trial an individual accused of a serious crime.” Sell, 539 U.S. at 180 , 123 S.Ct. 2174 . To determine whether a crime is “serious” enough to satisfy the first Sell factor, we first consider the likely Sentencing Guidelines range applicable to the defendant and then consider other relevant factors. Hernandez-Vasquez, 513 F.3d at 919 .

The government calculates Gillen-water’s likely Guidelines range as 33 to 41 months. That is lower than in, for example, Her nandez-Vasquez, where we noted that the government’s interest in prosecuting a defendant charged with illegally reentering the United States after removal could satisfy the first Sell factor. See id. at 911-12, 919 (identifying likely Guidelines range of 92 to 115 months). But the offense conduct alleged in this case is nonetheless serious enough to establish an important governmental interest in Gillen-water’s prosecution. Gillenwater is accused of making lurid and distressing threats against a bevy of government officials and employees. He allegedly threatened to, among other things, choke, rape, and kill people who serve our country. The threats allegedly continued for over a year, escalating in volume and violence. And the district court found that Gillenwa-ter “evidenced a possible intent and ability to carry out th[e] threats.” ER 314. In prosecuting him, the government is seeking “to protect through application of the criminal law the basic human need for security.” Sell, 539 U.S. at 180 , 123 S.Ct. 2174 . Indeed, the government is seeking to protect, the very integrity of our system of government.

Even when a defendant is charged with a serious crime, “[sjpecial circumstances may lessen the importance” of the government’s interest in prosecuting him. Id. Gillenwater contends that three such special circumstances are at play here. First, Gillenwater contends that the government has a lessened interest in prosecuting him because his “illness and resulting proclivities towards making threats might make him subject to” civil commitment. Appellant’s Br. 17. Eligibility for civil commitment may “diminish the risks that ordinarily attach to freeing without punishment one who has committed a serious crime,” Sell, 539 U.S. at 180 , 123 S.Ct. 2174 , thereby lessening the government’s interest in prosecuting him, Ruiz-Gaxiola, 623 F.3d at 694 . But there is nothing in the record to suggest that Gil-lenwater is eligible for civil commitment. The district court did not determine whether Gillenwater would meet the civil commitment criteria, and none of the experts who evaluated him took a position on that issue.

Second, Gillenwater contends that the government has a lessened interest in prosecuting him because he has been in custody for about 32 months. “[T]he possibility that the defendant has already been confined for a significant amount of time,” as Sell explained, “affects, but does not totally undermine, the strength of the need for prosecution.” 539 U.S. at 180 , 123 S.Ct. 2174 . Here, the government maintains a strong interest in prosecuting Gillenwater. For one thing, Gillenwater is charged with not one but three offenses. If convicted, he may face sentences that run consecutively. See 18 U.S.C. § 3584 . If convicted, Gillenwater may also face su pervised release, see 18 U.S.C. § 3588 , which several of our sister circuits have found significant in evaluating the government’s showing on the first Sell factor, see, e.g., United States v. Gutierrez, 704 F.3d 442, 451 (5th Cir.2013); United States v. Nicklas, 623 F.3d 1175, 1179 (8th Cir.2010); United States v. Bush, 585 F.3d 806, 815 (4th Cir.2009). Supervised release would help ensure that Gillenwater does not return to making threats when released into the public. And the monitoring that accompanies supervised release may be especially valuable here because Gillenwater allegedly persisted in making threats despite law enforcement intervention. After all, the government alleges that Gillenwater returned to sending threatening emails two days after federal agents warned Gillenwater that doing so could land him behind bars. What is more, the government alleges that Gillen-water sent a threatening postcard while in custody. Finally, as explained by one of our sister circuits, “the fact of a conviction would create certain limitations on [Gillen-water’s] subsequent activities, such as [his] ability to obtain and own firearms, see 18 U.S.C. § 922 (d)(1), (g)(1), which may be particularly important where, as here, [Gil-lenwater] is charged with making threats against federal [officials and employees].” Bush, 585 F.3d at 815 .

Third, Gillenwater contends that he “was inclined to commit the subject offenses at least in part because of his mental condition and, thus, such disorder should render it less important to criminally prosecute [him].” Appellant’s Br. 18. As we recognized in Ruiz-Gaxiola , the fact that a defendant’s mental disorder contributed to his offense may weaken the government’s interest in prosecuting him, but that will not always be the case. See 623 F.3d at 695. Here, the link between Gillenwater’s mental disorder and his charged crimes makes his prosecution all the more important. The government’s expert testified that without treatment, Gillenwater is likely to continue to act on his delusions. And the district court found that he “is at substantial risk to engage in the type of violent conduct which is the subject of the criminal proceeding against him if his mental condition is left untreated.” 9/24/13 Order 9.

B

Turning to the second Sell factor, the government must make a two-part showing to establish that involuntary medication will significantly further the important governmental interests at stake. First, the government must show that “administration of the drugs is substantially likely to render the defendant competent to stand trial.” Sell, 539 U.S. at 181 , 123 S.Ct. 2174 . And second, it must show that “administration of the drugs is substantially unlikely to have side effects that will interfere significantly with the defendant’s ability to assist counsel in conducting a trial defense.” Id.

Gillenwater does not challenge the district court’s determination that administration of haloperidol decanoate is substantially unlikely to have side effects that will interfere significantly with his ability to assist his counsel in his defense. He contends only that the district court clearly erred in determining that administration of the drug is substantially likely to render him competent to stand trial.

We cannot agree. The government’s expert, Dr. Lucking, testified that haloperi-dol decanoate is substantially likely to decrease Gillenwater’s delusional beliefs “to the extent where he will then be able to interact with and work with his attorney to plan a rational, logical defense.” ER 150. Dr. Lucking has a robust knowledge base, and he presented a strong case for the effectiveness of involuntary medication in rendering Gillenwater competent to stand trial. Dr. Lucking has been board certified in psychiatry and neurology for over 30 years. He has served as a staff psychiatrist at the Federal Medical Center in Butner, North Carolina, where he evaluated Gillenwater, for over 15 years. Based on his clinical experience, Dr. Lucking testified that defendants with psychotic disorders often attain competency when treated with antipsychotic medication. He also presented several recent studies indicating that more than 70% of persons suffering from delusional disorder saw an improvement in their symptoms when treated with antipsychotic medication. While those studies were not double-blind or placebo-controlled, both Dr. Lucking and Gillenwa-ter’s expert, Dr. Cloninger, testified that it would be very difficult, if not impossible, to conduct double-blind or placebo-controlled studies of the effectiveness of involuntarily medicating persons suffering from delusional disorder.

To support his position, Gillenwater points to Ruiz-Gaxiola , a case in which Dr. Cloninger testified and we held that the district court clearly erred in finding that the government satisfied the second Sell factor. See 623 F.3d at 695-701. As we noted in Ruiz-Gaxiola , Dr. Cloninger is an experienced psychiatrist with strong credentials. Id. at 690, 699-700 . But this case is not Ruiz-Gaxiola . Unlike in Ruiz-Gaxiola , Dr. Cloninger did not meet with or examine the defendant in this case before drawing his conclusions. See id. at 690 . Dr. Cloninger also copied and pasted substantial parts of his report in this case from the report he prepared in Ruiz-Gaxiola , including an erroneous reference to deportation. And most importantly, Dr. Cloninger relied exclusively on older studies. Dr. Lucking explained that prior to the 1990s, the commonly held psychiatric opinion was that delusional disorder rarely responded to treatment with antipsychotic medication. But Dr. Lucking relied on more recent studies indicating that the older negative view was mistaken. On the other side of the scale, Dr. Lucking is more experienced than the government’s psychiatrist in Ruiz-Gaxiola. See id. at 689-90, 699-700. And as described above, he made a strong case for involuntary medication, whereas the government’s psychiatrist in Ruiz-Gaxiola pointed to only one study supporting his treatment plan. See id. at 697-98 .

In sum, we cannot conclude that the district court clearly erred in accepting the testimony of an experienced expert who examined the defendant and showed a greater awareness and understanding of recent advances in the treatment of delusional disorder.

C

We next consider the third Sell factor — that involuntary medication is necessary to further the important governmental interests at stake. To satisfy the third Sell factor, the government must show that “any alternative, less intrusive treatments are unlikely to achieve substantially the same results.” Sell, 539 U.S. at 181 , 123 S.Ct. 2174 . And the district court must also “consider less intrusive means for administering the drugs ... before considering more intrusive methods.” Id.

Gillenwater does not contend that the district court failed to consider less intrusive means for administering the halo-peridol decanoate. He contends only that the district court clearly erred in rejecting an alternative, less intrusive treatment option, namely the voluntary psychotherapy suggested by Dr. Cloninger.

Dr. Cloninger opined that “the only appropriate medical treatment” for Gillenwa-ter was voluntary psychotherapy. ER 268. His treatment option required providing “a non-adversarial therapeutic milieu in which [Gillenwater] can recognize that he is respected and valued so as to allow the natural recovery process to take place unimpeded.” Id. But Dr. Lucking testified that voluntary psychotherapy was bound to fail because Gillenwater does not recognize that he suffers from a mental disorder and does not believe that he needs treatment. And Dr. Lucking also testified that, because of his mental disorder, Gillenwater “does not trust anybody” — a major roadblock to voluntary psychotherapy, which requires “the formation of a therapeutic alliance” built on trust between the psychiatrist and his patient. Id. at 180 , 123 S.Ct. 2174 .

The district court did not clearly err in crediting Dr. Lucking and rejecting Dr. Cloninger’s alternative treatment option. Dr. Cloninger provided conflicting opinions in his report and testimony on whether the type of voluntary psychotherapy he envisioned could be effectively performed at the federal facility in which Gillenwater is confined. He also testified that his treatment option is “just not going to be very effective” if Gillenwater refuses it. Id. at 379, 123 S.Ct. 2174 . And Gillenwater “clearly reserved the right to refuse any treatment he did not like.” 9/24/13 Order 11 n.5. As we have previously recognized, a district court does not clearly err in rejecting voluntary psychotherapy as an alternative, less intrusive treatment option where, as here, it is far from clear both that it can be effectively performed within the constraints of the prison environment and that the defendant will engage in it. See Ruiz-Gaxiola, 623 F.3d at 702-03 .

D

Finally, we assess the fourth Sell factor. To establish that administration of the drug is medically appropriate, the government must show that involuntary medication is in “the patient’s best medical interest in light of his medical condition.” Sell, 539 U.S. at 181 , 123 S.Ct. 2174 . In assessing whether the government has made that showing, we consider “the long-term medical interests of the individual rather than the short-term institutional interests of the justice system.” Ruiz-Gaxiola, 623 F.3d at 703 .

Gillenwater contends that the district court clearly erred in determining that involuntary treatment with haloperi-dol decanoate is in his best medical interest because the potential harms of that treatment outweigh the potential benefits. On the potential harms side, long-term treatment with haloperidol decanoate can result in severe side effects, including tar-dive dyskinesia, tardive dystonia, and tar-dive akathisia. But both Dr. Lucking and Dr. Cloninger testified that such severe side effects rarely emerge when an individual is treated with haloperidol decanoate for only a short time, which is what Dr. Lucking proposed for Gillenwater. Moreover, Dr. Lucking testified that Gillenwa-ter had no medical conditions and was not taking any medications that would predispose him to severe side effects. He also testified that Gillenwater would be given a small test dose to ensure that “there are no major adverse clinical effects or side effects” and that any side effects from that test dose could be reversed. ER 148. And Dr. Lucking testified that Gillenwater would be closely monitored throughout the treatment and that his treatment would be adjusted if any side effects did emerge.

On the potential benefits side, Dr. Lucking testified that involuntary treatment with haloperidol decanoate was not only substantially likely to render Gillenwater competent to stand trial but that it could also open the door to other treatment options. In particular, Dr. Lucking testified that it could facilitate treatment with newer antipsychotic medications that have a lower incidence of severe side effects with long-term administration but that can be administered only voluntarily. And Dr. Lucking testified that Gillenwater, unlike, for example, the defendant in Ruiz-Gaxio-la, is “experiencing significant distress as a result of his condition.” 623 F.3d at 705. He explained that Gillenwater “believes that he is being harassed, persecuted, and threatened with bodily harm, even death, ... based upon his possession of certain information which he believes that others do not want released into the community.” ER 145. “When you believe that people are out to kill you because of information that you have and that you have to run for your life, that people are hunting you down, that they’re plotting against you and conspiring against you and you always have to watch your back,” Dr. Lucking elaborated, “that is a very painful experience” for which treatment “is in the best interests of the individual.” Id. at 399-400. And based on their observations of Gillenwater, both Chief Judge Peterson and Judge Suko found that he is “deeply disturbed by his delusions and could not lead a fulfilling life in the absence of some improvement in his condition.” Id. at 334; 9/24/13 Order 12.

In sum, we cannot conclude that the district court clearly erred in determining that involuntary medication is in Gillenwa-ter’s best medical interest when the potential harms and benefits of the treatment are viewed against the seriousness of his condition.

For the reasons discussed above, we AFFIRM the district court’s September 24, 2013 involuntary medication order (Case No. 13-30284). We DISMISS as moot Gillenwater’s appeal of the district court’s November 19, 2012 involuntary medication order (Case No. 12-30379).

. In Hatper, the Supreme Court held that the government may “treat a prison inmate who has a serious mental illness with antipsychotic drugs against his will, if the inmate is dangerous to himself or others and the treatment is in the inmate's medical interest.” 494 U.S. at 227 , 110 S.Ct. 1028 . We have since explained that "the district court, in an ordinary case, should refrain from proceeding with the Sell inquiry before examining” whether a defendant may be medicated involuntarily under Harper. United States v. Hernandez-Vasquez, 513 F.3d 908, 914 (9th Cir.2008). As the district court noted in authorizing involuntary medication under Sell , Dr. Lucking concluded that Gillenwater did not meet the Harper criteria, and the government therefore moved for a Sell order. Under those circumstances, "we cannot fault the district court for honoring the parties' agreement to proceed directly to the Sell inquiry.” Id. at 915. . The September 24, 2013 order, which incorporates the November 19, 2012 order, is the involuntary medication order currently in effect. We therefore consider Gillenwater’s arguments as applied to the September 24, 2013 order. Because the November 19, 2012 order is no longer operative, we dismiss as moot Gillenwater’s appeal of it.