OBJECTIVES: To determine whether (i) motivational enhancement therapy (MET) + cognitive behaviour therapy (CBT) compared with usual care, (ii) MET compared with usual care, (iii) or MET + CBT compared with MET was more effective in improving glycaemic control when delivered by general nurses with additional training in these techniques. DESIGN: A three-arm parallel randomised controlled trial as the gold standard design to test the effectiveness of psychological treatments. SETTING: The recruiting centres were diabetes clinics in seven acute trusts in south-east London and Greater Manchester. PARTICIPANTS: Adults (18-65 years) with a confirmed diagnosis of type 1 diabetes for a minimum duration of 2 years and a current glycated (or glycosylated) haemoglobin (HbA1c) value between 8.2% and 15.0%. INTERVENTIONS: The control arm consisted of usual diabetes care which varied between the hospitals, but constituted at least three monthly appointments to diabetes clinic. The two treatments arms consisted of usual care with MET and usual care with MET + CBT. MAIN OUTCOME MEASURES: The primary outcome was HbA1c at 12 months from randomisation. Secondary outcome measures were 1-year costs measured by the Client Service Receipt Inventory at baseline, 6 months and 12 months; quality of life-years [quality-adjusted life-years (QALYs)] measured by the SF-36 (Short Form-36 Health Survey Questionnaire) and EQ-5D (European Quality of Life-5 Dimensions) at baseline and 12 months. RESULTS: One thousand six hundred and fifty-nine people with type 1 diabetes were screened and 344 were randomised to MET + CBT (n = 106), MET (n = 117) and to usual care (n = 121). The 12-month follow-up rate for HbA1c was 88% (n = 305). The adjusted mean 12-month HbA1c was 0.45% lower in those treated with MET + CBT [95% confidence interval (CI) 0.16% to 0.79%, p = 0.008] than for usual care; 0.16% lower in those treated with MET (95% CI 0.20% to 0.51%, p = 0.38) than for usual care; and 0.30% lower with MET + CBT than with MET (95% CI -0.07% to 0.66%, p = 0.11). The higher the HbA1c, and the younger the participant at baseline, the greater was the reduction in HbA1c. The interventions had no effect on secondary outcomes such as depression and quality of life. The economic evaluation was inconclusive. Both interventions were associated with increased health care costs than for usual care alone. There was no significant difference in social costs. Cost effectiveness ratios, up to one year, varied considerably according to whether QALY estimates were based on EQ-5D or SF-36 and whether imputed or complete data were used in the analyses. CONCLUSIONS: A combination of MET and CBT may be useful for patients with persistent sub-optimal diabetic control. MET alone appears less effective than usual care. Economic evaluation was inconclusive. TRIAL REGISTRATION: Current Controlled Trials ISRCTN77044517.