This review assesses whether motivational interviewing (MI)-based interventions are effective at improving health behaviours in adults with diabetes. Electronic databases were searched for articles that investigated the use of MI and diabetes self-management between 1966 and March 2010. In total, 464 titles were found; after duplicates were removed, 112 studies remained. Of these, 24 abstracts were identified as potentially relevant. No studies were excluded on the basis of their methodology. By use of a data extraction sheet, eight studies were identified as relevant once full articles were examined. Positive results in health-behaviour improvement were reported in four studies. These were: reduced smoking, improved blood glucose, improved diet and weight management. Studies varied in quality; four were methodologically weak owing to small sample sizes, lack of clarity of scoring measurement tools and/or limited use of valid measurements, and reported inclusion/exclusion criteria. In conclusion, MI cannot be recommended as an evidenced-based approach to diabetes self-management.