There is evidence for a team-based approach in the management of chronic disease in primary health care. However, the standard of care is variable, probably reflecting the limited organisational capacity of health services to provide the necessary structured and organised care for this group of patients. This study aimed to evaluate the impact of a structured intervention involving non-GP staff in GP practices on the quality of care for patients with diabetes or cardiovascular disease. A cluster randomised trial was undertaken across 60 GP practices. The intervention was implemented in 30 practices with staff and patients interviewed at baseline and at 12-15 months follow up. The change in team roles was evaluated using a questionnaire completed by practice staff. The quality of care was evaluated using the Patient Assessment of Chronic Illness Care questionnaire. We found that although the team roles of staff improved in the intervention practices and there were significant differences between practices, there was no significant difference between those in the intervention and control groups in patient-assessed quality of care after adjusting for baseline-level score and covariates at the 12-month follow up. Practice team roles were not significantly associated with change in Patient Assessment of Chronic Illness Care scores. Patients with multiple conditions were more likely to assess their quality of care to be better. Thus, although previous research has shown a cross-sectional association between team work and quality of care, we were unable to replicate these findings in the present study. These results may be indicative of insufficient time for organisational change to result in improved patient-assessed quality of care, or because non-GP staff roles were not sufficiently focussed on the aspects of care assessed. The findings provide important information for researchers when designing similar studies.