Diagnostic phototesting, including the determination of the minimal erythema dose (MED), is a useful procedure to detect abnormal sensitivity to UV radiation. We aimed to estimate the reference limits (RLs) of the MED in a reasonably large reference sample of white individuals. Skin phototypes and MED values for broadband UVB and for UVA were determined in 461 white subjects. When appropriate, the 95% reference intervals, including the 0.025 fractile and 0.975 fractile, were computed for the MED-UVB reference values (by means of parametric methods) and the MED-UVA reference values (by means of nonparametric methods). MED data were also converted to standard erythema doses (SEDs). As described elsewhere we observed a considerable overlap of MED values for all skin phototypes and confirmed that age and sex do not substantially influence the MED. The lower RLs observed for MED-UVB were 33 mJ cm⁻² (0.5 SEDs) and for MED-UVA 12.6 mJ cm⁻² (1.2 SEDs). The MED and SED findings from this investigation may serve as reference data for white individuals and give support to the clinician in differentiating between normal and pathologically abnormal photosensitivity. Although the MED data given here are limited to the phototest device used in the present study, the SED results establish comparability between our data and phototest results obtained from laboratories using different UV sources.