Lifestyle programmes delivered by academic nurse-counsellors in a primary care setting lowered blood pressure (BP) among at-risk patients. We examined whether the programmes could be successfully implemented by nurses employed in Australian general practices. In a randomised controlled trial, 212 of 591 eligible 20-75-year olds with hypertension, Type II diabetes or coronary disease from seven practices volunteered. Patients were randomised to: a 'Low' group with one face-to-face individual counselling session, then monthly telephone contacts for 1 year (n=69); a 'High' group with individual face-to-face counselling up to 1 h monthly for 1 year (n=74) or a control group receiving usual care only (n=69), and were evaluated at baseline and 12 and 18 months later; 164 individuals completed the study. Patients' usual doctors continued to prescribe in all groups. Changes in 24 h ambulatory BP did not differ significantly between groups at 12 months (Low, -2+/-2/1+/-1 mmHg; High, +4+/-2/1+/- 1 mmHg; usual care, +1+/-2/1+/-1 mmHg) or 18 months (-2+/-2/2+/-1 mmHg; -4+/-2/3+/-2 mmHg; -1+/-2/2+/- 1 mmHg, respectively). Antihypertensive drugs prescribed decreased by 12 months in 33% of the High, 5% of the Low and 13% of the control groups (P=0.008) and by 36, 7 and 16% at 18 months (P=0.018). After 18 months, targets for BP control were not met in about 60% of patients and almost 50% had clinic BP above 140/90 mmHg. Year-long interaction with nurse-counsellors may influence longer-term antihypertensive drug prescription, possibly by improving compliance. Suboptimal BP control suggests that continuing physician education on BP targets is needed.