FELDMAN, Justice.

We have accepted review of this case to end any confusion over the reach of our decision in Dutra v. Industrial Commission, 135 Ariz. 59, 659 P.2d 18 (1983). A worker with a rated 30 percent permanent partial loss of use of his left leg claims that our holding in Dutra entitles him to benefits for a scheduled 100 percent disability because he can no longer perform his prior employment. He received an award based on the 30 percent disablement. The court of appeals affirmed, 148 Ariz. 575, 716 P.2d 32. We granted review pursuant to Rule 23, Ariz.R.Civ.App.Proc., 17A A.R.S. We have jurisdiction pursuant to Ariz. Const. art. 6, § 5( 3) and A.R.S. § 12-120.24.

FACTS

Antonio Gomez' (Gomez) job as a truck driver and freight hauler required that he lift heavy weights and perform an extensive amount of climbing and bending. In September, 1971 Gomez injured his left knee in a work-related accident. His workers' compensation claim was accepted and surgery was performed on the knee. Ultimately, he was paid $3,750 for a 15 percent permanent partial loss of use of his left leg pursuant to A.R.S. § 23-1044(B)(21) and his file was closed.

In the Spring of 1979 he began having more problems with his left knee. His original claim was reopened on the recommendation of his attending physician and surgery was again performed in August, 1979. It was noted that there was marked degeneration and osteoarthritis of the knee as a result of the original injury. Gomez' knee continued to deteriorate and he underwent more surgery in the Fall of 1981.

In July, 1982 Gomez' medical condition was determined to be stationary and in August, 1982 his file was closed with findings that he had a 30 percent permanent partial loss of use of his left leg as determined by the AMA Guides to the Evaluation of Permanent Impairment. The doctors were in agreement that he could no longer perform most of the tasks required in his old employment and that he could not return to his former work.

Gomez appealed his award, arguing that a degenerative back problem was caused by the change in gait associated with the injury, thus converting his injury from a scheduled to an unscheduled one. Because it was undisputed that he could no longer perform his job, he also argued, in the alternative, that under Dutra, supra, he was entitled to scheduled benefits for a 100 percent loss of use, the same amount he would have received if his leg had been severed. A.R.S. § 23-1044(B)(15).

After weighing conflicting medical evidence, the administrative law judge (ALJ) determined that Gomez' back problems were unrelated to the knee injury. Therefore, accepting the doctors' rating, he made a scheduled award of 30 percent permanent partial loss of use of the left leg. As to the Dutra issue, the ALJ found that all the doctors had agreed upon a 30 percent impairment, that there was no range on the percentage of impairment, and that it would be "sheer arbitrary speculation" for him to increase the award. The court of appeals affirmed, holding that there was no abuse of discretion by the ALJ. The court of appeals also raised serious questions regarding Dutra's reach and effect. Gomez v. Industrial Commission, 148 Ariz. 575, 716 P.2d 32 (App. 1985). We accepted review.

DISCUSSION

We agree with the court of appeals that substantial evidence supported the ALJ's determination that Gomez' back problems were unrelated to his knee injury and that he suffered a scheduled permanent partial loss of use to his left leg; therefore there was no abuse of discretion by the ALJ on these questions. (At 148 Ariz. at 576, 716 P.2d at 33). Thus, we must decide whether the proper factors were applied to determine the percentage of partial loss of use.

When a worker is injured there are two basic classifications of injury — scheduled and unscheduled. Subsection (B) of A.R.S. § 23-1044 sets forth the schedule of compensation to be paid for the permanent partial disability from loss of certain bodily members or functions. Subsections (B)(1) through (B)(20) provide a fixed schedule of payments for total loss of various digits, limbs and bodily functions. Each subsection provides for compensation to be paid for a specific number of months at a fixed percentage of the employee's average monthly wage. Subsection (B)(21) provides that where there has been a "partial loss of use" of a particular digit or limb the injured worker is entitled to receive scheduled compensation equal to "that proportion of the number of months in the . . . schedule provided for the complete loss of use . . . which the partial loss of use thereof bears to the total loss of use. . . ." Thus, in all cases where there has been a "partial loss of use" it is necessary to determine the percentage of impairment that has been sustained. The statute does not indicate how partial loss of use is determined and does not specify any factor to be considered in determining the percentage of impairment which will determine the degree of "partial loss of use." It does state in § 23-1044(H) that an injury which is not an unscheduled injury must be considered as scheduled and compensated under subsection (B) "regardless of its actual effect on the injured employee's earning capacity." Therefore, effect on an injured employee's earning capacity may not be considered in determining the amount of "partial loss of use" for a scheduled injury under subsection (B)(21).

Subsection (C) of A.R.S. § 23-1044 requires that for unscheduled injuries, compensation be based on loss of earning capacity rather than a fixed schedule. Subsection (D) specifies the factors which must be considered when determining the amount of lost earning capacity in unscheduled injury cases. These factors include, "among other things", previous disability, the occupational history of the employee, the nature and extent of the physical disability, the type of work for which the injured employee will be fit in the future, subsequent wages, and the age of the employee. There is no express statutory prohibition in subsection (D) or elsewhere on the use of these factors in the determination of "partial loss of use" for scheduled injury under subsection (B)(21). Obviously, one of them — nature and extent of physical disability — must be considered in determining percentage of impairment for scheduled injuries as well as in determining loss of earning capacity for unscheduled injuries. On the other hand, earning capacity cannot be considered in a scheduled injury case, this being expressly prohibited by subsection (H).

The statute leaves unresolved the problem of determining just what factors should be considered when assessing the percentage of "partial loss of use" under subsection (B)(21) for scheduled injuries involving less than total loss. The Industrial Commission has adopted a regulation (A.C.R.R. R4-13-113(D)) to be applied in making these assessments. It provides that the:

Rating of the percentage of functional impairment [loss of use] shall be in accordance with the standards for the evaluation of permanent impairment as published by the American Medical Association in `Guides to the Evaluation of Permanent Impairment'.

In Adams v. Industrial Commission, 113 Ariz. 294, 552 P.2d 764 (1976), we examined the use of the AMA Guides under the cited Commission rule. We concluded that the Guides were not an "all-encompassing rating of all possible . . . impairments." Id. at 295, 552 P.2d at 765. We further concluded that in situations "where the Guide does not apply, it should not be used to limit the percentage of impairment to that portion which is rated therein." Id. We stated further:

We hold that Rule 13(D) must be so interpreted that the AMA Guides apply exclusively to the evaluation of permanent impairment to the extent that the AMA Guides cover the specific impairment and the percentage thereof. Beyond this Rule 13(D) has no application.

Id. at 295-96, 552 P.2d 765-66 (emphasis supplied). Adams held that although the AMA Guides provided for only an 11 percent impairment, petitioner was entitled to a rating of a 30 percent impairment for total loss of hearing. This assessment was based on standards promulgated by the U.S. Labor Department because those standards more accurately reflected the actual amount of functional loss. Id. at 295, 552 P.2d at 765. Adams did not require us to determine what other factors may be considered in evidence if the AMA Guides are found to be inapplicable in rating a particular impairment.

Dutra, supra, was a case in which the AMA Guides were of little value in measuring the percentage of impairment of use of petitioner's arm because petitioner had no loss of movement of the arm, but rather had a loss of the arm's strength for lifting. The court of appeals held in Dutra that, notwithstanding the Industrial Commission regulation, "an impairment not ratable under the AMA Guides may be rated by other criteria". 135 Ariz. 86, 87, 659 P.2d 45, 46 (App. 1982), citing Adams v. Industrial Commission, supra (emphasis supplied). The ability to perform a particular job was not to be considered. Id. at 88, 659 P.2d at 47. In making the use of other factors discretionary, the court of appeals incorrectly construed Adams. When the Guides are inapplicable, Adams requires that other criteria be used to determine "just compensation". 113 Ariz. at 295, 552 P.2d at 765. Despite this error, in Dutra the court of appeals approved a rating in excess of that which would have been authorized under the AMA Guides. The ALJ had based the increased rating upon the doctor's abstract assessment of Dutra's impairment. Our opinion in Dutra went one step further; we held that the effect of the injury on a worker's ability to perform his or her job could also be considered as a factor in determining the extent of impairment. Dutra, 135 Ariz. at 62, 659 P.2d at 21.

In the present case the court of appeals states that Dutra is contrary to Smith v. Industrial Commission, 69 Ariz. 399, 214 P.2d 797 (1950), because Dutra allows the Commission to consider "the occupation of the injured employee" — a factor mentioned in subsection (D) — while Smith holds that subsection (D) is applicable only to an unscheduled injury. ( 148 Ariz. at 577, 716 P.2d at 34.) Smith did not hold that the factors mentioned in subsection (D) cannot be used to determine the percentage of "partial loss of use" under subsection (B). As noted above, one of the subsection (D) factors ("nature and extent of physical injury") must be used under subsection (B). Smith stands only for the proposition that a subsection (B) scheduled injury cannot be compensated as if it were a subsection (C) unscheduled injury. Our opinion in Dutra does not negate or affect that principle.

The court of appeals feared that Dutra permitted the ALJ to consider the effect of the injury on the employee's earning capacity, contrary to subsection (H) of the statutes. ( 148 Ariz. at 578, 716 P.2d at 35.) This is not what Dutra held. Dutra holds only that the effect of the injury on the employee's ability to perform the work which he was doing at the time of the injury should be considered in determining the degree of physical impairment. Dutra, 135 Ariz. at 61, 659 P.2d at 20. It does not state that the injury's effect on earning capacity should be taken into consideration. There is a significant difference. The ability to do a particular job may be affected without having any effect on earning capacity. Effect on earning capacity is not a factor to be considered in a scheduled injury case, See § 23-1044(H).

Dutra did hold that "inability to perform the claimant's particular job at the time of his injury must be considered in determining the extent of the workman's disability." 135 Ariz. at 61, 659 P.2d at 20 (emphasis supplied). This holding was made in a factual context in which the AMA Guides were unable to accurately measure the actual partial loss of use. We did not state that inability to perform the job was equivalent to total loss of use of the injured member. Our statutory scheme could not tolerate such a result because it would erase the legislative distinction between scheduled and unscheduled injuries. See Smith, supra.

The Commission has adopted the AMA Guides for rating the percentage of impairment when there has been a partial loss of use rather than the complete loss of a member. See A.R.S. § 23-1044(B)(21). When they are applicable and "truly reflect the claimant's loss", they may be used as the sole indicator or factor to be considered in fixing the percentage of impaired function. Adams v. Industrial Commission, 113 Ariz. at 295, 552 P.2d at 765. Where the ALJ finds that the Guides do not provide a fair, accurate measure of the degree of impairment, he or she must turn to other factors. Id. Any relevant factors, except those prohibited by statute (such as loss of earning capacity) may be considered. Effect on job performance is one such factor. Dutra, supra. Evidence regarding such factors may come from experts, from the literature, lay witnesses or any other competent source that would assist the ALJ in determining the actual percentage of partial loss of use. Use of these factors fulfills the statutory mandate.

CONCLUSION

Dutra is not, therefore, the radical departure feared by the court of appeals. No revolution in workers' compensation law has occurred. It is obviously legally impossible to allow blind application of the Commission's regulation to cases where it is an inaccurate predictor. The Commission has no power to adopt guidelines and then require their application in all cases, even when they fail to provide an accurate measure of the loss. Adams, supra. Dutra merely follows previous case law in holding that in situations where the AMA Guides fail to accurately measure loss of use, the award determining the degree of partial loss must be based on other evidence. It adds one relevant factor not considered before in such cases. That factor, like all other relevant evidence, should be considered in determining the percentage of impairment in a scheduled injury case. Effect of the injury on earning capacity should not be considered. Once the percentage of impairment is determined, with or without the AMA Guides, then the scheduled award is computed according to the statute just as it has always been.

We turn now to apply these principles to the case before us. All doctors agreed that the AMA Guides provided an accurate measure of the degree of Gomez' impairment. Therefore, the ALJ did not err in basing his award on the rating provided by the Guides. The award is affirmed. The opinion of the court of appeals is approved as modified.

HOLOHAN, C.J., GORDON, V.C.J., and HAYS and CAMERON, JJ., concur.