Contact allergy to methyldibromo glutaronitrile (MDBGN), often combined with phenoxyethanol (PE) (e.g., Euxyl K 400), increased throughout the 1990s in Europe. Consequently, in 2003, the European Commission banned its use in leave-on products, where its use concentration was considered too high and the non-sensitizing use concentration as yet unknown. The 2 objectives of the study are (a) to find a maximum non-eliciting concentration in a leave-on product in MDBGN/PE-sensitized patients, which could possibly also be considered safe regarding induction and (b) to find the best patch test concentration for MDBGN. We, therefore, performed a use-related test (ROAT) in patients sensitized to MDBGN/PE (n = 39) with 3 concentrations of MDBGN/PE (50, 100 and 250 p.p.m. MDBGN, respectively). A subset of these patients (n = 24) was later patch-tested with various concentrations (0.1, 0.2, 0.3 and 0.5% MDBGN, respectively). 15 patients (38%, 95% confidence interval (CI) = 23-55%) had a negative and 24 (62%; 95% CI = 45-77%) a positive overall repeated open application test (ROAT) result. 13 reacted to the lowest (50 p.p.m.), 8 to the middle (100 p.p.m.) and 3 to the highest concentration (250 p.p.m.) only. In those 13 reacting to the lowest ROAT concentration, dermatitis developed within a few days (1-7). The strength of the initial and the confirmatory patch test result, respectively, and the outcome of the ROAT were positively associated. Of the 24 patients with a use and confirmatory patch test, 15 reacted to 0.1% MDBGN, 16 to 0.2%, 17 to 0.3% and 22 to 0.5%. With the patch test concentration of 0.5%, the number of ROAT-negative patients but patch-test-positive patients increases considerably, particularly due to + reactions. A maximum sensitivity of 94% (95% CI = 70-100%) is reached with a patch test concentration of 0.2%, and is not further improved by increasing the concentration. However, the specificity decreases dramatically from 88 (95% CI = 47-100%) with 0.2% to a mere 12.5% (95% CI = 0-53%) with 0.5%. It can be concluded (a) that for MDBGN 0.2% is very likely the best patch test concentration and (b) that 50 p.p.m. in a leave-on product can elicit contact dermatitis in sensitized persons. We were, therefore, unable to find a safe, still microbicidal, concentration for leave-on products. By contrast, with other contact allergens, dose-response use tests may be able to identify a non-eliciting concentration, which could give valuable clues to a non-inducing (i.e., safe) concentration in products.