Abstract Background It is now well established that the integration of mental health care into primary care is one of the most effective ways of reducing the substantial treatment gap for mental disorders which exists in most low- and middle-income countries. This study set out to determine whether a Mental Health Gap Action Programme (mhGAP) training and supervision package could be contextualised and implemented within the existing health care system in five districts in Southern Malawi. In addition, the study assessed the feasibility of holding community awareness events and establishing peer support groups in each district to further improve the access of the population to evidence-based mental health care. Methods A lead training team of experienced Malawian mental health professionals was appointed and mhGAP training materials were contextualised for use in Malawi. The lead team delivered a 4-day training package to district mental health teams in five districts covering three core conditions: psychosis, moderate-severe depression, and alcohol and substance use disorders. District mental health teams then delivered a 2-day training package and provided monthly supervision for 3 months to 500 non-specialist healthcare workers. Paired sample t-tests were used to compare knowledge, confidence and attitude scores before and immediately after training, and after 6 months in two districts. Case detection rates measured pre- and post-training in the pilot district were compared using Wilcoxon Rank Sum Test. Community awareness events were held and peer support groups were established in each of the five districts. The acceptability of the package was assessed through focus group discussions involving specialist and non-specialist healthcare workers, users and carers. Results Non-specialist healthcare workers’ knowledge and confidence scores significantly increased immediately after training in comparison to pre-training. These scores were maintained at 6 months. However, no statistically significant change in attitude scores was detected. Case detection rates increased immediately after the training in comparison to pre-training. Responses from focus group discussion participants illustrated the programme’s acceptability. Conclusions This study demonstrated that, with minimal additional funding and working within existing structures, an mhGAP based training at primary and secondary health care levels is feasible in Southern Malawi.