Photodynamic therapy (PDT) with topical 5-aminolevulinic acid (ALA) is increasingly employed for skin cancer, yet ALA dosing is crude. Using iontophoresis, we developed a rapid and quantifiable system for topical ALA delivery, with measurement of subsequent PpIX fluorescence and phototoxicity. ALA was iontophoresed from a 2% solution into upper inner arm skin of 13 healthy volunteers. Six doses of ALA were delivered with a series of charges varying from 3-120 milliCoulombs (mC); four additional doses were given with a charge of 60 mC. Five hours post-iontophoresis, sites were irradiated with broad-band yellow-red light, the series of six ALA doses receiving 100 J/cm2, while the four identical doses received 6.25, 12.5, 25, and 50 J/cm2, respectively. Resultant erythema was measured by reflectance spectroscopy. The time course of PpIX fluorescence was ALA-dose-dependent. With charge < or = 24 mC, PpIX fluorescence peaked at 3 h and returned to zero at 9-10 h, whereas charges > 24 mC had a sustained peak at 5-10 h, falling to zero by 24 h. Pre-irradiation, PpIX fluorescence correlated with ALA dose (r = 1.0). PpIX fluorescence fell immediately post-irradiation (p < 0.0001); recovery levels at 3 h correlated with ALA dose (p < 0.0001). Delayed erythema correlated with ALA dose and irradiation dose (p < 0.0001, p < 0.01, respectively). Both PpIX fluorescence intensity pre-irradiation and fall in PpIX fluorescence post-irradiation correlated with erythema (r = 0.98). Hence, PpIX synthesis is ALA-dose-dependent, and phototoxicity can be predicted from ALA dose, irradiation dose, and photobleaching of PpIX. This reproducible system allows accurate dosimetry in topical PDT and facilitates study of ALA metabolism.