A photopatch test series consisting of 27 substances was tested in 81 patients with suspected photosensitivity disorders. Irradiation was performed using the following light sources: TL-K 40W/09 bulbs (UVA'; lambda max. at 355 nm), UVASUN 5000 (UVA''; lambda max. at 375 nm), and TL 20 W/12 bulbs (UVB; lambda max. at 315 nm). One day after applying 4 sets of the test substances (D1), one test series each was exposed to 10 J/cm2 of UVA' or UVA'', or a combination of 40 mJ/cm2 UVB and 10 J/cm2 UVA''; the fourth series was left nonirradiated (control). Photopatch test reactions (PPTR) were defined as positive if there was at least an indurated erythema on D3 or later between D3 and D21 (late-onset reactions) exclusively at the irradiated, but not at the control site. At least one positive PPTR was found in 35 patients (43%), 7 of whom exhibited late-onset reactions only; in 2 cases the tests could not be read because of skin irritation. On D3, there were 44 positive reactions with UVA', only 10 of which were also demonstrable with UVA''. Twenty-one late-onset PPTR were found with UVA' and 17 with UVA''. Late-onset reactions elicited by UVA' or UVA'' mostly were concordant; divergent positive or negative results were found only in a few cases. Compared with the results obtained with UVA' or UVA'' alone, combined irradiation with UVB and UVA'' occasionally led to divergent positive as well as negative results. When photopatch test results are interpreted, it should be considered that testing with different UVA sources may yield divergent results.