Background: Bath-PUVA treatment, originally established in Scandinavia, offers several advantages over oral PUVA and has become increasingly popular in recent years. Outside Scandinavia 8-methoxypsoralen (8-MOP) is the prevailing photosensitizer for this PUVA modality and is used arbitrarily in a wide range of concentrations. Up to the present, data are lacking on the impact of 8-MOP bath-water concentration on UVA dosimetry. Objective: We investigated the influence of increasing 8-MOP bath-water concentrations on photosensitivity in bath-PUVA treatment. Methods: Fifteen healthy volunteers without abnormal photosensitivity or recent exposure to ultraviolet radiation were included in an intraindividually controlled comparison study. In all volunteers the minimal phototoxic dose (MPD) was determined on the volar side of their forearms after immersion for 20 minutes in 4 different 8-MOP bath-water concentrations (0.5, 1, 2.5, and 5 mg/L). The correlation between 8-MOP concentration and photosensitivity (defined as the reciprocal value of the MPD) was analyzed by linear regression analysis. In addition, the time course of erythema formation and the UVA dose-erythema response curve was assessed for each psoralen concentration. Results: The median MPD and the 25%-75% interquartile were 5.7 J/cm² (5.7-8), 4 J/cm² (4-5.7), 2.8 J/cm² (2.8-5.7), and 2 J/cm² (2-2.8) at an 8-MOP concentration of 0.5, 1, 2.5, and 5 mg/L, respectively. Linear regression analysis revealed a significant correlation between 8-MOP bath-water concentration and photosensitivity (r = 0.98; P = .019). Bath-PUVA-induced erythema peaked after a median time interval of 3 days, with a range of 2 to 4 days. The slope of the UVA dose-erythema response curve was similar for all psoralen concentrations. Conclusion: UVA dose requirements in bath-PUVA treatment decrease linearly with increasing 8-MOP concentrations. A single MPD assessment at 72 hours after the UVA exposure is inappropriate for accurate determination of the patients' photosensitivity. The hazard of wrong UVA dosimetry is comparable at all psoralen concentrations. (J Am Acad Dermatol 2001;44:638-42.)