Background Limited research has investigated how to promote physical activity in people with type 2 diabetes. This study evaluated physical activity counseling over 12 months in people with type 2 diabetes. Design Participants were given standard exercise information and randomly assigned to receive physical activity counseling or not. Data were collected from September 2000 through to September 2002 and analyzed from October 2002 to February 2003. Setting/participants Diabetes outpatient clinic. Seventy inactive people with type 2 diabetes. Intervention Physical activity counseling, based on the transtheoretical model, combined motivational theory and cognitive behavioral strategies into an individualized intervention to promote physical activity. Consultations were delivered at baseline and 6 months, with phone calls at 1 and 3 months post-consultation. Main outcome measures Changes from baseline at 12 months in physical activity (7-day recall and accelerometer), stages and processes of exercise behavior change. Results Between-group differences were recorded in physical activity (recall and accelerometer) at 12 months (p <0.01). Experimental participants significantly increased total activity (median difference, 115 minutes; 95% confidence interval [CI]=73–150 minutes). Control participants recorded no significant change (median difference, -15 minutes; 95% CI=-53–13 minutes). The accelerometer experimental participants recorded no significant change (mean difference, 416,632 counts; 95% CI=-27,743, 1,051,007 counts/week), while control participants recorded a significant decrease (mean difference, -669,061 counts; 95% CI=-1,292,285, -45,837 counts/week). At 12 months, more experimental participants compared to controls were in active stages of behavior(6-month ?2=26.4, p <0.01; 12-month ?2=19.9, p <0.01, respectively). Between-group differences were recorded at 12 months for the frequency of using all processes (p <0.01), except dramatic relief and stimulus control. Conclusions Physical activity counseling was effective for promoting physical activity over 12 months in people with type 2 diabetes.