The aim of the study was to examine whether the minority of practices not qualifying for payment for structured diabetes care programmes differ systematically from those that do. Information was collected for all Leicestershire general practices on practice size, population structure, deprivation indices, diabetes related admissions over two years and number of insulin treated patients on the district register. The 21 practices not offering structured diabetes care had a median list size of 3204, compared to 6340 for the other 124 practices (P < 0.001). Jarman and Townsend scores were higher for these practices and estimated prevalence of diabetes was 29% higher (95% CI: 26-32%). Crude admission rates were significantly higher in those practices not offering structured care. However rates adjusted for diabetes prevalence were similar (39.3 vs 39.2 per 100 insulin treated diabetics per year, P = 0.9). These results suggest that some practices face specific problems related both to small practice size and higher prevalence. If these issues are not addressed, inequalities in access to diabetes care between practice populations will persist. There is no evidence that the provision of structured care is associated with lower admission rates in this district. However more information, particularly in relation to prevalence of diabetes, is needed in order to accurately quantify this relationship. Variations in prevalence between practices should be adjusted for in any comparison of admission rates or spurious conclusions may be drawn.