In a previous publication from our group, phototesting based on a single exposure to a divergent UVB beam with radially decreasing irradiance values was suggested. The aim of the present study was to evaluate technical, practical and biological aspects of the suggested method in normal subjects. Twenty healthy volunteers were provoked on the back with both a collimated beam (four fixed doses, in circular areas with a diameter of 1.5 cm) and the divergent beam (a continuous, radially attenuating dose spectrum covering an area with a diameter of 4.5 cm). Eleven of the subjects were subjected to double provocation with the divergent beam. Assessment was carried out at 6 and 24 h after exposure by measuring the diameter of the reactions both visually and by mapping the skin blood flow change with laser Doppler perfusion imaging (LDPI). Minimal erythemal dose (MED) was determined for both the collimated and the divergent provocation. The reaction diameters were used to decide MED by combination to a mm for mm mapped dose spectrum of the divergent beam profile. Dose-response curves were plotted using the quantitative response data of the LDPI-images against the corresponding dosimetry data. No systematic difference could be proven between LDPI and visual diameters and a 95% confidence interval for the mean difference was calculated to (-0.8, 2.0). Slightly greater diameters were found at the visual assessment performed at 6 h compared to 24 h (95% confidence interval (-0.1, 2.8)). Double provocation showed a good reproducibility both for the visual and the LDPI assessment (P<0.05). The divergent beam provocation allowed a more detailed discrimination of MED compared to the collimated beam provocation. The MED values determined with the divergent beam were, however, generally higher, especially in the lower range of MED values. Technical factors related to the beam divergence and the correct measurement of erythemal effective irradiance are believed to be the explanation for this phenomenon, which is thus correctable. In conclusion, the results from this study support our belief that the phototesting protocol based on a divergent beam constitutes a good opportunity for improved phototesting, since MED and dose-response characteristics may be extracted in more detail from a single UV exposure.