BackgroundSelf-prescribing among doctors is common, but no longitudinal studies have documented this issue. We studied the self-prescribing behaviour among young Norwegian physicians and the predictors of self-prescribing.MethodsWe conducted a nationwide, prospective and longitudinal study following young Norwegian physicians from internship through the subsequent nine years using three postal questionnaires. Chi-square tests and logistic regression models were applied.ResultsAbout 54% of the physicians in their fourth and ninth postgraduate years had self-prescribed medication at least once during the previous year. Among those who had used prescription medication during the previous year, about 90% had self-prescribed. Self-prescribing behaviour did not differ significantly between men and women, or according to the type of work at any time. The most frequently self-prescribed medications were antibiotics (71% 81%), contraceptives (24% 25%), analgesics (18% 21%), and hypnotics (9% 12%). Those who had needed treatment for mental problems self-prescribed hypnotics and sedatives to a greater extent than the others. Being male, having self-prescribed during internship, somatic complaints, mental distress, subjective health complaints, and not having sought help from a general practitioner were significant adjusted predictors of self-prescribing in the ninth postgraduate year. ConclusionsThe level of self-prescribing among young Norwegian physicians is relatively high, and this behaviour is established early in their professional lives. Although self-prescribing is acceptable in some situations, physicians should seek professional help for illness. Efforts to implement a more rational help-seeking behaviour should probably start in medical schools.