OBJECTIVES: To determine whether practice structure (for example, list size, number of staff) predicts team processes and whether practice structure and team process in turn predict team outcomes DESIGN: Observational study using postal questionnaires and medical note audit. Team process was assessed through a measure of "climate" which examines shared perceptions of organisational policies, practices, and procedures. SETTING: Primary care. SUBJECTS: Members of the primary health care team from 42 practices. MAIN OUTCOME MEASURES: Objective measures of quality of chronic disease management, patients' evaluations of practices, teams' self-reported ratings of effectiveness, and innovation. RESULTS: Team climate was better in singlehanded practices than in partnerships. Practices with longer booking intervals provided superior chronic disease management. Higher team climate scores were associated with superior clinical care in diabetes, more positive patient evaluations of practice and self-reported innovation and effectiveness. CONCLUSIONS: Although the conclusions are preliminary because of the limited sample size, the study suggests that there are important relationships between team structure, process, and outcome that may be of relevance to quality improvement initiatives in primary care. Possible causal mechanisms that might underlie these associations remain to be determined.