Aims Most patients with diabetes are treated in primary care (PC). We performed a systematic review to assess the effect of single and combined interventions on cardiovascular risk factors (CVRF) and HbA1c levels in patients with diabetes in PC settings. Methods We searched the MEDLINE database from January 1990 through October 2008. According to the Cochrane Effective Practice and Organization of Care Group (EPOC) criteria, (cluster-) randomised control studies and controlled before-and-after studies were selected and reviewed. Identified interventions were classified according to a modified EPOC intervention taxonomy. Results We included 68 studies. 45 studies evaluated the effect of any intervention on HbA1c. 17 studies presented a significant improvement in HbA1c. Nine out of 27 studies evaluating CVRFs (Cholesterol, Blood Pressure) and HbA1c showed a significant improvement in at least two of these factors. Audit and feedback on performance, clinical decision support systems, multi-professional teams and patient education seemed to be successful strategies. Conclusions The increasing evidence regarding the treatment of persons with chronic illnesses, summarised in the Chronic Care Model (CCM), is not reflected in most recent studies about diabetes treatment in primary care. Most interventions still seem only partly adapted to the CCM. The methodological quality of many studies is still poor and often the pivotal outcomes CVRFs and HbA1c are not appropriately addressed. As a consequence, the potential of PC in the care of patients with diabetes may still be underestimated.