Successful results have been reported for bath—psoralens plus ultraviolet A (PUVA) treatment of psoriasis with the use of either a low-sensitizing psoralen (methoxsalen) or a high-sensitizing psoralen (trioxsalen), but no evidence has been presented that phototoxicity would be a prerequisite for antipsoriatic activity of these treatments. To study therapeutic-to-phototoxic ratios, bath-PUVA was applied to psoriasis patients with the use of either a 0.2 mg/L solution of trioxsalen or a 0.4 mg/L solution of methoxsalen. The average minimal phototoxic ultraviolet A (UVA) dose (MPD) obtained after 15 minutes' bathing was 0.86 joule/cm² for trioxsalen and 9.76 joules/cm² for methoxsalen. In the first part of the study phototoxically equipotent treatment schedules were used, compensating the much lower phototoxic potential of methoxsalen by using about 10 times larger UVA doses compared to the doses used with trioxsalen. A healing rate of 71% ± 25% was recorded for trioxsalen and 63% ± 34% for methoxsalen treatment (mean ± SD). Both PUVA treatments decreased epidermal Langerhans cell counts to 10% to 20% of the control value. In the second part of the study, the much lower phototoxic potential of methoxsalen/was not compensated for. Instead, methoxsalen bath-PUVA was carried,'but with the use of UVA doses physically similar to those used in the trioxsalen therapy series. Surprisingly enough, even this very suberythemal methoxsalen therapy caused a significant healing effect (51% ± 10%). Langerhans cell depletion in methoxsalen-bathed areas (reduction to 53% of the control value) was much less than that caused in trioxsalen-bathed areas (reduction to 11% of the control value). Since Langerhans cells are considered to be important in the immunologic protection of the skin against malignancies, low UVA-dose methoxsalen bath-PUVA therapy might be an advantageous PUVA regimen.