Background. In high Andean areas, there is greater insulin sensitivity which may be a protective factor against complications in subjects with diabetes mellitus. Objective: Determine the association between altitude of residence and death from COVID-19 in deaths with diabetes mellitus in Peru during the first wave.. Methods: We carried out a cross-sectional analysis of deaths registered in the National Death System of Peru (SINADEF in Spanish). We selected Peruvians with diabetes mellitus identified by presenting the diagnosis in any of the six boxes on the certificate. The dependent variable was death from COVID-19 as the basic cause of death, located in last place among causes A, B, C and D according to the Pan American Health Organization. The independent variable was the altitude of residence, categorized as less than 1 500 m a.s.l, 1 500 to 2 499 m a.s.l. and greater than 2 500 m a.sl. Through a multilevel analysis by geographic region and using a Poisson regression, we obtained the risk ratios of death from COVID-19 according to the altitude of residence. We adjusted by individual and contextual variables. Results: We included 16 406 deaths with diabetes mellitus between March-December 2020. 34.3% died from Covid19 and 9.7% came from areas above 2 500 m. The proportion of deaths from COVID-19 of those with residence altitude above 2 500 m was 20% lower compared to residents below 1 500 m (RR: 0.80; 95% CI: 0.70 – 0.91; p<0.001), adjusted for individual and socioeconomic factors. Its influence is also shown as the altitude changes every 100, 250, 500 and 1000 m a.s.l., through multilevel analysis. Conclusion: A higher altitude of residence is associated with a lower proportion of deaths from COVID-19 in people with diabetes mellitus during the first wave in Peru. The study contribute to expanding knowledge of the effects of altitude with respect to mortality in people with diabetes mellitus in a context of a highly contagious and virulent infectious disease.