The consequences of a diagnosis of occupational contact dermatitis (OCD) were investigated using cases available from the voluntary surveillance scheme, EPIDERM. Cases of OCD reported from November 1994 to September 1995 were identified and sampled to give at least 100 cases of allergic, irritant and mixed OCD reported by consultant dermatologists (344 cases) and occupational physicians (377 cases). A questionnaire was sent to the reporting physician to elicit further information. 512 completed questionnaires were returned, of which 510 were eligible for analysis. Among cases reported by dermatologists (n = 286) and occupational physicians (n = 224), 7% (6.3%) had been unemployed and 16.8% (20.1%) had taken sick leave. 3 factors independently predicted time off work in a logistic regression analysis: age OR = 1.25 (95% CI, 1.05-1.49), allergic dermatitis OR = 1.77 (95% CI, 1.13-2.79) and medicolegal assessment OR = 4.42 (95% CI, 2.20-8.89). Overall, 15.7% did not improve clinically between the first and last visit. Those who did not improve had been exposed to the agent for longer (mean 7.6 years) than those who did (5.3 years) (p = 0.09). In patients <or= 45 years, those reported to be atopic failed to improve (25.4%) more often than those not atopic (13.4%) (p = 0.04). The substantial numbers (21%) with time off work and with persistent dermatitis suggest that OCD continues to have a significant impact on workers and their employers.