Background: Despite the evidence that physical activity (PA) can prevent type 2 diabetes mellitus (T2DM), limited research investigated the level of PA among diabetes and nondiabetes in Myanmar, where there is the escalating prevalence of diabetes recently. We investigated PA as modified the risk of diabetes, in a case-control study. Methods: We conducted a case-control study which included 150 cases and 150 controls age 25-74 years (Mean age 43.3±14.7 years) among the cases and (55.1±10.9 years) among the controls, both sex and residence in Yangon. Cases were newly diagnosed with T2DM within six months before data collection, with laboratory-confirmed fasting blood glucose level ≥126mg/dl. Controls were community residents, without diabetes, confirmed with a laboratory test. The IPAQ-S was used to assess the PA level. Multiple logistic regression analysis was applied in STATA 15, using the interaction terms for age and PA, adjusting age, sex and BMI. Results: In comparison to controls, cases were older and having less PA knowledge. The levels of vigorous PA were mean 254.9±standard deviations (SD) 845.6 METs min.wk−1 among controls, 73.06±392.1 cases, moderate PA 631.5±1240.8 METs min.wk−1 among controls and 1050.9±1601.6 cases and walking PA 569.8±1060 METs min.wk−1 among controls and 777.4 ±1249 cases, respectively. The multiple logistic regression analysis showed adjusted odds ratios (aOR) 3.84, 95% confidence interval (CI) 1.18-12.42 (P<0.05) for those aged 40 and older, with moderate PA and aOR 18.01, Cl 6.45-50.26 (P<0.001) for those aged 40 and older, with low PA. Comparing the strength of association, the risk of T2DM among people aged 40 and older with moderate PA is lower than age over 40 with low PA. Conclusion: PA lessened the risk of T2DM, posted by increasing age. The findings highlighted the importance of promoting PA to reduce the T2DM prevalence in the context of Yangon, Myanmar, a low-and middle-income Asian country.