Background: Primary care teams have large numbers of patients with psychological problems in their sole care, and many organisations will wish to design educational opportunities to improve and maintain their knowledge and skills. However, very few clinical tutors have clear guidance from the evidence-based literature to assist them. Previous studies have shown that simple methods such as guideline distribution rarely change professional behaviour or alter patient outcomes, and relevant reviews have rarely highlighted the key components which make interventions effective. This qualitative review extracted data from studies on educational interventions in primary mental health care in order to evaluate effectiveness and give guidance. Methods: Systematic searching of named databases using specified keywords and time period, for both original articles and reviews, focusing on randomised controlled trials (RCTs). Data extraction and quality reviews were carried out as well as exploration of common factors associated with significant outcomes. Findings: Eighteen papers in total fitted the search criteria. Key features of significant outcomes were multifaceted educational interventions including practice based components in the intervention, such as practice visits, or use of practitioners' own patient cases and data, This finding fits with educational principles known to assist in adult learning. Recommendations: The impact of educational interventions which support constructive learning from personal data are likely to be more effective than courses working with theory alone or 'imaginary' patients, even when these are well designed and carried out. Using examples and data from practice appear to be crucial for showing clinicians their potential needs for improvement, and to engage them in an active cycle of learning. Educational interventions need to be rigorously designed, and all components evaluated, in order to know what makes them effective.