IntroductionIn recent years, the government of Bangladesh has encouraged private sector involvement in producing mid-level health cadres including Medical Assistants (MAs). The number of MAs produced has increased significantly. We assessed students’ characteristics, educational services, competencies and perceived attitudes towards health service delivery in rural areas.MethodsWe used a mixed method approach using quantitative (questionnaire survey) and qualitative (key informant interviews and roundtable discussion) methods. Altogether, five public schools with 238 students and 30 private schools with 732 students were included. Statistical analyses were performed using STATA v-12. Qualitative data were analyzed thematically.FindingsThe majority of the students in both public (66%) and private medical assistant training schools (MATS) (61%) were from rural backgrounds. They spent the majority of their time in classroom learning (public 45% versus private 42%) and the written essay exam was the common form of a students’ performance assessment. Compared with students of public MATS, students of private MATS were more confident in different aspects of educational areas, including managing emerging health needs (P<0.001); evidence-based practice (P=0.002); critical thinking and problem solving (P=0.02), and use of IT/computer skills (P<0.001). Students were aware of not having adequate facilities in rural areas (public 71%, private 65%), but they perceived working in rural areas will offer several benefits, including use of learnt skills; friendly rural people; and opportunities for real-life problem solving, etc.ConclusionThis study provides a current picture of MATS students’ characteristics, educational services, competencies and perception towards working in rural areas. The MA students in both private and public sectors showed a greater level of willingness to serve in rural health facilities. The results are promising to improve health service delivery, particularly in rural and hard-to-reach areas of Bangladesh.