The aim of this study was to estimate the risk of vulvo-vaginal candidiasis among the users of acitretin. The incidence rate ratio of vulvo-vaginal candidiasis was estimated in a cohort of acitretin users by using prescription sequence analysis. Study subjects were 196 women between 15 and 45 years of age, exposed to acitretin in 1990. Prescriptions of drugs for treatment of vulvo-vaginal candidiasis were used as proxy for the occurrence of vulvo-vaginal candidiasis. Validation of the proxy drug was performed by record linkage of pharmacy and medical records. In addition to the risk estimation by incidence rate ratios a case-crossover study with patient-stratified logistic regression analysis was performed to control for distortion due to individual differences in contribution of exposure time. The positive predictive value of the proxy drug for vulvo-vaginal candidiasis ranged between 57 and 100%, the sensitivity was 87% and the specificity estimated 99%. The crude incidence rate ratio for vulvo-vaginal candidiasis following acitretin exposure was 2.8 (CI95%: 1.1-7.1). The pooled Mantel-Haenszel incidence rate ratio was 3.3 (CI95%: 1.1-9.6) after stratification for accumulated level of exposure. Patient-stratified analysis on the subgroup of cases (n = 15) revealed an odds ratio of 6.5 (CI95%: 2.3-18.2). This study strongly suggests that the higher occurrence of vulvo-vaginal candidiasis during acitretin exposure is attributable to the drug.