Reading of doubtful patch test reactions can be improved by comparing them to defined negative, allergic and irritant controls. For the latter, an irritant patch test is needed that gives sufficiently reproducible results. In our double-blind, placebo-controlled, randomized study, we have analysed the synchronous reproducibility of patch tests with 0%, 0.0625%, 0.125%, 0.25%, 0.5% and 1.0% sodium lauryl sulfate (SLS). Tests and readings were done according to ESCD guidelines. 139 patients (75 women, 64 men) 18 to 77 years old were tested. The % of positive patients as well as the % of reproducible positive reactions increased with rising concentrations of SLS to a maximum of approximately 90% and 85%, respectively, with 1% SLS. Time courses of reactions were also related to the concentration of SLS: with 1% SLS, plateau-type time courses were most common (65%), followed by decrescendo- (23%) and crescendo-type reactions (12%). With 1% SLS, very similar mild to moderate reactions were obtained on days 2 and 3. The reaction threshold and reproducibility of tests were not related to age and sex of patients or to their own assessment of skin irritability. We conclude that 1% SLS is appropriate for an irritant patch test that gives sufficiently reproducible results.