OBJECTIVE: To evaluate, on a patient level, the effect of a "physician-level intervention" that successfully reduced the incidence of anticholinergic antidepressant prescribing. DESIGN: Cross-sectional surveys with questionnaires sent before and after intervention. SETTING: Additional study in an RCT to reduce the prescribing of highly anticholinergic antidepressants in the elderly in the South Holland Islands. PARTICIPANTS: Elderly patients (age 60-95 years) who used antidepressants in 1995 and 1996 in our research area according to a health insurance prescription database. MAIN OUTCOME MEASURES: Prevalence of adverse events related to antidepressant use, severity of depression and quality of life were compared in users living in the intervention and control areas. RESULTS: Prior to our intervention we sent 2,359 questionnaires of which we could use 827 (35%) for analysis. At baseline, there were no statistically significant differences between the intervention and control areas. After the intervention 3,375 questionnaires were sent, of which 939 (28%) could be used. The occurrence of "dry mouth" and "coughing" and the "amount of pain" were lower in the intervention area compared to the control area (p < 0.05). CONCLUSION: We found no indications that adverse events, severity of depression or quality of life were changed in an unfavourable direction, when comparing patients inside and outside the intervention area.