Education and psychosocial factors are widely believed to contribute to diabetes control. However, there is no systematic way of measuring or recording these factors, or for transfer of information to future care-givers. We developed a 41-question questionnaire to measure the skills and contributory factors to self-management in diabetes, and examined the relationship between these factors and glycaemic control. The questionnaire was assessed for construct validity, and for content validity by independent diabetes health professionals, and easily completed by 128 insulin-treated diabetic patients (mean age 42 +/- 13 (SD) yr, diabetes duration 17 +/- 9 yr, HbA, 10.5 +/- 1.8%). Principle Components Factor analysis revealed nine factors explaining 63% of variance (Eigen values 1.2-5.8). Five of these were significantly related to HbA1. These were: practical self-management skills, p < 0.001; emotional adjustment, p < 0.002; perceived goals, p < 0.01; perceived self-efficacy, p < 0.005; and costs-benefit analysis, p < 0.01. Multiple regression revealed that these five predictor variables (factors) explain 21% of variance in HbA1, combined r = 0.46. The questionnaire also distinguished between patients with good and poor control (HbA1 8.2 vs 13.4%, p < 0.001), p for differences in question responses, 0.014-0.001. Four factors (lifestyle, influence of other people, diet concerns, and weight concerns) were unrelated to HbA1. In conclusion, we have developed a single, patient-completed questionnaire of high content and construct validity, which identifies alterable factors (in contrast to fixed demographic variables) which may be specifically targeted in individual patients on the one hand to improve glycaemic control and, on the other, to enable more satisfactory adjustment to life with diabetes. It may be used in day-to-day practice to identify patient needs, and their barriers to effective self-management, and to transfer this information to future care-givers.