The use of antidepressants (ADs) has escalated and prompted considerable debate. Many depressed patients go unrecognised or under-treated and the area of indication of the new ADs is widening. The aim of this study was to analyse (i). the variation in general practitioners' prescribing of ADs by comparing with prescribing of other drug groups and (ii). whether the general prescribing behaviour, practice activity and demography are associated with the AD prescribing.Analysis of AD prescribing patterns among 174 general practices (93.5%) in the County of Funen, Denmark. Age- and sex-standardised 1-year incidences and prevalences of AD prescribing for patients listed were calculated using individual prescription data from Odense University Pharmacoepidemiologic Database. Data about health services and practice demography were obtained from the Health Insurance Register. The variation in AD 1-year prevalence was compared with other drug groups by a variation index (90%/10% percentile). Univariate linear regression analysis was used to examine associations between practice characteristics and prescribing.The 1-year prevalence of AD prescribing varied sixfold, no more than the prevalence of five other drug groups. Practices with high yearly: general prescribing prevalence, mean number of drugs per medicated patient, number of surgery consultations/100 patients and counsellings/100 surgery consultations showed the highest yearly prevalence of AD prescribing. Single-handed practices had higher AD prescribing rates than partnerships. The relative use of selective serotonin re-uptake inhibitors and other new ADs showed only little variation (10% and 90% percentiles as close as 66-86%), but practices with high 1-year prevalence and incidence most often chose the new ADs.Analysis of inter-practice variation showed no extraordinary quality problems with regard to AD prescribing, but does not exclude that there might be problems. The general prescribing pattern of the general practitioners seems essential to their attitude to AD prescribing. The relationship between counselling and prescribing was a feature specific to ADs and deserves further investigation. Quality indicators are needed to understand differences in AD prescribing, and studies based on prescription data have to be supplemented with individual clinical data.