The study objective was to assess sick-leave among women investigated in a general population survey of alcohol dependence/abuse (ADA). A total of 399 women, selected by stratified random sampling, were interviewed and diagnosed according to DSM-III-R. Data on sick-leave were obtained by linkage with the Social Insurance records. The study found that women with ADA but without other psychiatric disorders had an increased number of annual sick-leave spells--1.82 compared with 1.47 in the reference population-whereas the mean duration was similar. Women with ADA and other psychiatric disorders had 2.38 annual spells, but also considerably longer spells (mean 16.54 days vs 9.70). Socio-economic differences were large, with the less privileged groups having both more and longer spells. Stepwise multiple regression showed that both ADA and other psychiatric disorders contributed to high sick-leave incidence and duration, as, to a lesser extent, did low education and low social group (the last variable only affected duration of the spells). The frequency of disability pension/long sickness spells was higher in women with ADA (odds ratio of 2.95). We concluded that there is a strong association between ADA and sick-leave, which increases considerably in the presence of additional psychiatric disorders.