On June 20, 1960, Fred Montana injured his back lifting a heavy steel blade onto a truck. A coworker holding one end of the blade dropped it without warning, and Montana, suddenly holding almost the entire load, "felt something tight in . . . [his back].” The injury, an acute lumbosacral strain, resulted in temporary total disability and partial permanent disability. Petitioner paid Montana temporary compensation of $928.57 at the rate of $65 a week before the Industrial Accident Commission made its award. The referee determined that 75 per cent of the disability should be apportioned to a preexisting back condition, that Montana’s earning capacity was less than minimum for compensation purposes, and that the award should therefore be based on average weekly earnings of $20 a week. Computed at this rate, Montana would be entitled to $285.72 as temporary compensation and $440 for permanent disability. Under this computation the insurance carrier had overpaid Montana $243.57. The commission held, however, that the evidence did not require apportionment of the disability under Labor Code section 4663 and that Montana’s "earning capacity was maximum under Section 4453(d) of the Labor Code.” Accordingly, the commission awarded him $4,620 as permanent indemnity and approved the temporary compensation paid.

Montana testified that he had never had back trouble before the injury. No evidence contradicted this testimony, and several doctors described his former condition as “asymptomatic.” Nevertheless, the medical experts agreed that even before the injury he had an unstable lumbosacral spine with some arthritis and that he should not have been lifting heavy *593 objects. When asked what advice he would have given Montana before the accident, the independent medical examiner testified: “I don’t know—I couldn’t tell him not to lift, because I know he would. We look at a number of things. I would not advise him to give up his work, if this is all he knows how to do, because some of these guys do get by indefinitely.” (Italics added.) Describing Montana’s preaccident condition, he stated: “There was some increase in the lumbosacral angle with mild scoliosis. His pelvis was tipped forward. He had some degenerative changes which were not marked. I mean, there was nothing unusual about the x-rays for a man of this build . . . and age. They are not too far away from normal, about what you would expect with his build and age.” (Italics added.) Another doctor stated in his report: “There is no evidence of pre-existing symptomatic bone or joint disease and no history of previous low back injury or disability. The osteoarthritie changes visualized are no more in evidence in the area of symptoms than they are in other areas of his spine. At most they delayed his functional recovery.”

Whether a disability results in whole or in part from “the normal progress of a preexisting disease” (Industrial Indem. Co. v. Industrial Acc. Com., 95 Cal.App.2d 443, 450 [ 213 P.2d 11 ]) or represents a fully compensable lighting up or aggravation of a preexisting condition is a question for the commission to determine, and its award will not be annulled if there is substantial evidence to support it. (Colonial Ins. Co. v. Industrial Acc. Com., 29 Cal.2d 79, 83-84 [ 172 P.2d 884 ] ; Tanenbaum v. Industrial Acc. Com., 4 Cal.2d 615, 617 [ 52 P.2d 215 ].) In view of the foregoing testimony, the commission could reasonably conclude that the preexisting condition did not call for apportionment. (See Idaho-Maryland Mines Corp. v. Industrial Acc. Com., 104 Cal.App.2d 567 [ 232 P.2d 11 ].)

The more difficult question is whether the commission correctly determined Montana’s earning capacity under subdivision (d) of section 4453 of the Labor Code 1 in computing *594 temporary and permanent disability compensation. Other subdivisions of that section (Lab. Code, §4453, subd. (a), (b), (c)) set forth formulae for computing average weekly earnings that in turn are made the basis for the two types of award. (Lab. Code, §§ 4653-4655 ; 4658-4662.) When an employee is steadily employed at a full-time job his earning capacity is determined by an appropriate formula (see West v. Industrial Acc. Com., 79 Cal.App.2d 711, 722 [ 180 P.2d 972 ]). When the employment is for less than 30 hours a week or when a formula “cannot reasonably and fairly be applied’’ the commission must make its own estimate of weekly earning capacity at the time of the injury. (Lab. Code, §4453, subd. (d).) The purpose of this provision is to equalize for compensation purposes the position of the full-time, regularly employed worker whose earning capacity is merely a multiple of his daily wage and that of the worker whose wage at the time of injury may be aberrant or otherwise a distorted basis for estimating true earning power. It would hardly be consistent with that purpose to foreclose a worker from a maximum temporary or permanent award simply because a brief recession had forced him to work sporadically or at a low wage. Nor in making a permanent disability award would it be consistent with the purpose of the statute to base a finding of maximum earning capacity solely on a high wage, ignoring irregular employment and low income over a long period of time.

An estimate of earning capacity is a prediction of what an employee’s earnings would have been had he not been injured. Earning capacity, for the purposes of a temporary award, however, may differ from earning capacity for the purposes of a permanent award. In the former case the prediction of earnings need only be made for the duration of the temporary disability. In the latter the prediction is more complex because the compensation is for loss of earning power over a long span of time. Thus an applicant’s earning capacity could be maximum for a temporary award and mini *595 mum for a permanent award or the reverse. Evidence sufficient to sustain a maximum temporary award might not sustain a maximum permanent award. In making an award for temporary disability, the commission will ordinarily be concerned with whether an applicant would have continued working at a given wage for the duration of the disability. In making a permanent award, long-term earning history is a reliable guide in predicting earning capacity, although in a variety of fact situations earning history alone may be misleading. With regard to both awards all facts relevant and helpful to making the estimate must be considered. (Colonial Mut. Comp. Ins. Co. v. Industrial Acc. Com., 47 Cal.App.2d 487, 490-492 [ 118 P.2d 361 ] ; Aetna Life Ins. Co. v. Industrial Acc. Com., 130 Cal.App. 488, 491-492 [ 20 P.2d 372 ] ; see Southern Bell Tel. & Tel. Co. v. Bell (Fla.) 116 So.2d 617 , 620-621 ; Vanney v. Alaska Packers Assn., 12 Alaska 284 , 290-291 ; Larson, The Law of Workmen’s Compensation, § 57.21 at pp. 4-7.) The applicant’s ability to work, his age and health, his willingness and opportunities to work, his skill and education, the general condition of the labor market, and employment opportunities for persons similarly situated are all relevant. (See West v. Industrial Acc. Com., supra at p. 722 ; Aetna Life Ins. Co. v. Industrial Acc. Com., supra at pp. 491-492.) In weighing such facts, the commission may make use of “ ‘its general knowledge as a basis of reasonable forecast. ’ ” (Latour v. Producers Dairy, Inc., 102 N.H. 5 [ 148 A.2d 655, 657 ] ; compare Russell v. Southeastern Util. Service Co., 230 Miss. 272 [ 92 So.2d 544, 547 ].) In weighing the evidence relevant to earning capacity the commission has the same range of discretion that it has in apportioning injuries between industrial and nonindustrial causes. (See e.g., Aetna Life Ins. Co. v. Industrial Acc. Com., supra at p. 493.) It must, however, “have evidence that will at least demonstrate the reasonableness of the determination made.” (Davis v. Industrial Com. of Arizona, 82 Ariz. 173 [ 309 P.2d 793, 795 ].)

Montana was 50 years old when he was injured, and he had established a pattern of intermittent work that apparently extended over five years. He worked as a laborer, usually on construction projects, was generally paid at a high hourly rate, and took any employment offered to him through his union. He held four jobs during the 15 months preceding his injury. Two of them lasted only three days. During the winter of 1959-1960 he worked 15 to 17 days as a brick *596 tender, earning between $104 and $125 a week. At the time of his injury, he had been employed for two months on a flood control project on which he earned $887.20. Total earnings over the 15-month period amounted to between $1,200 and $1,300.

In measuring earning capacity, the commission must give “due consideration to actual earnings from all sources and employments.” (Lab. Code, § 4453, subd. (d).) The evidence herein of a long record of irregular employment and low earnings would seem, therefore, to require a finding of minimum earning capacity in making the permanent award unless there were other relevant facts tending to support a contrary conclusion. In this regard, the commission stated that “the duration of . . . [Montana’s] work was indefinite,” and that he was “a permanent employee working irregularly due to business conditions.” The only sense in which the job was indefinite, however, was that no one could say for certain how many weeks it would last. The fact that Montana was hired only for a particular project that was completed within three months of the date of his injury indicates that the job was contemplated to be of limited duration. The findings that he was a “permanent employee” and that his sporadic employment was caused by “business conditions” are ambiguous. The commission could not mean that Montana normally held a steady job or that his employment at the time of injury was “permanent.” It probably meant to distinguish the case of a person who is a constant competitor on the labor market from one who by choice enters and leaves that market periodically. (See Larson, supra, at §§ 60.20, 60.21, 60.22, pp. 72-78 ; compare Campbell v. Industrial Acc. Com., 95 Cal.App.2d 570 [ 213 P.2d 395 ] and West v. Industrial Acc. Com., supra, with O’Hearne v. Maryland Cas. Co., 177 F.2d 979 .) In this sense, the finding that Montana was a “permanent employee” is essentially no more than a finding of his willingness to work. Willingness to work, however, can be meaningful in predicting what an applicant’s earnings would have been only if considered in the light of his employment opportunities. With regard to such opportunities the commission stated that “business conditions” caused Montana’s low and unstable earning history. If the business conditions causing temporary low earnings are no more than a short recession, the earnings will most likely not represent earning capacity for purposes of a permanent award. Long-run business conditions, however, necessarily have a bearing *597 on ability to earn. The commission did not specify the nature of the business conditions referred to. Had it heard evidence tending to show that Montana’s five lean years were unusual for him, that the construction industry in the area anticipated a period of prosperity, that men with similar skills, ages, physical characteristics, and earning histories were becoming steadily employed, it could properly have discounted Montana’s earning history. In the absence of such evidence, however, the commission failed to give due consideration to actual earnings as required by section 4453.

The commission had substantial evidence, however, to support its award for temporary disability. Montana worked for two months on the project, which lasted for an additional three months following the injury, and he would probably have continued thereon had he not been disabled.

The award of temporary disability compensation is affirmed. The award of permanent disability compensation is annulled, and the cause is remanded for further proceedings not inconsistent with this opinion.

Gibson, C. J., Schauer, J., McComb, J., Peters, J., White, J., and Dooling, J., concurred.

Petitioner’s application for a rehearing was denied June 4, 1962.

The statute read at the time the award was made: