Gout is a common disease. Although effective treatments are available for gout, there is some disagreement as to how they should be used. To study prescription patterns in gout, we conducted a questionnaire survey among 2520 rheumatologists. Seven hundred and fifty completed questionnaires were returned over a two-month period. Among respondents, 35.4% worked in a private office, 21% in a hospital and 43.6% in both. The most widely prescribed treatments in acute gout attack were colchicine alone (63%), colchicine with a nonsteroidal antiinflammatory drug (NSAID) (31.7%) and NSAID alone (5.2%), with significant variations according to the type of practice. Mean duration of treatment in acute gout was 18 +/- 16.8 days (range, 3-180 days). Mean time interval between the attack and initiation of therapy with a xanthine oxidase inhibitor was 21.6 +/- 17.2 days (range 0-180); here also, significant variations were seen according to the type of practice. Concomitant symptomatic therapy was prescribed in 97.3% of cases, for a mean duration of 54 +/- 55.4 days (range 2-365). Thirty per cent of responders never prescribed uricosuric agents. The estimated rate of occurrence of treatment-induced attacks increased with the reported interval between the attack and initiation of urate-lowering therapy. Our data demonstrate that French rheumatologists have widely diverging views on how to treat gout. Whether a waiting period is needed between an acute attack and initiation of urate-lowering therapy, and how long this period should be, are unsettled issues that deserve to be studied.