Introduction: Diabetes caused 1.5 million deaths in 2012. High blood glucose resulted in an additional 2.2 million deaths, by increasing the risk of cardiovascular and other diseases. Non-compliance with treatment diabetes mellitus patients can potentially increase morbidity, mortality and treatment costs. The aim of this study is to improve medication adherence and controlled random blood glucose in the hope reduce mortality and morbidiry of diabetes. Method: Quasi experimental in Posbindu RW 003 and 006 Kedaung Kaliangke with samples of diabetes mellitus type 2, the sampling method was total sampling. The independent variable was an intervention that was carried out in the form of giving a medicine box, calendar, individual education and taking medication. The dependent variable was the change in RBG. Result: 20 respondents who met the inclusion criteria. The mean RBG before intervention was 253 (201 -591) mg / dL and the mean RBG after intervention was 182.5 (130 -220) mg / dL. Wilcoxon Test showed that there were significant differences in RBG levels between before and after the intervention (p-value: 0.001). Conclusion: Knowledge enhancement through individual education has proven to be effective in increasing adherence to type 2 diabetes mellitus treatment and controlled RBG.