OBJECTIVES: To assess trends in prescriber activity related to patient smoking status and ACORN socio-economic status. The findings will be used to identify variables for subsequent regression analysis of the relationship between prescriber activity and consultation length. METHODS: The DIN-LINK database was used to obtain detailed data on prescriber activity. DIN-LINK is a longitudinal patient database comprising over 3.8 million electronic patient records, populated with data from General Practices across Great Britain. This analysis was based on patients consulting with a GP in the year to August 2004. Patients were stratified by smoking status and by ACORN socio-economic classification (CACI Ltd., www.caci.co.uk/acorn); measures of prescriber activity were examined, including the number of prescriptions per patient/per consultation, and the number of consultations per patient. The unpaired student t-test was used to test for statistical significant differences between groups. RESULTS: The cohort comprised 467,446 patients, of which 339,275 had a record of smoking status and 415,418 had their ACORN socio-economic status recorded. Among the sub-group with smoking status recorded, significant differences were found between current smokers, non-smokers, and ex-smokers in mean consultations per patient, prescriptions per patient, and prescriptions per consultation. Although significant results were found, no trends were apparent. The only non-significant comparisons were in mean prescriptions per consultation between current smokers and ex-smokers, and in mean consultations per patient between current smokers and non-smokers. When considering patients' socio-economic status, those in the lowest group (“Hard Pressed”) consulted more frequently and were issued significantly more prescriptions per head and significantly more prescriptions per consultation compared to the highest ACORN group (“Wealthy Achievers”). CONCLUSIONS: We found statistically significant differences in prescriber activity with respect to patients' recorded smoking and socio-economic status. Our findings have highlighted the potential of these two factors as suitable variables for regression analysis planned for 2005.