OBJECTIVE: To estimate the mortality attributable to diabetes mellitus (DM) as recorded by Peru's Ministry of Health and its association with the human development index (HDI). METHODS: This was an ecological study based on a secondary analysis of death records belonging to the Ministry of Health for the period from 2005 to 2014. A death was considered attributable to DM if the underlying cause of death given in the death record was DM. Mortality attributable to DM has been presented descriptively and in terms of geospatial analyses, and Spearman's rho was used to test for an association between the difference in the mortality attributable to DM (between 2005-2006 and 2013-2014) and the HDI in Peru's various departments. RESULTS: In the 10-year period under evaluation, 25 074 records listed DM as the underlying cause of death. The mortality rate attributable to DM per 100 000 -population increased from 5.7 in 2005 to 9.5 in 2014. This accounted for 2.7% of the deaths recorded during the study period: 3.5% in coastal areas, 1.4% in the highlands, and 2.5% in the rainforest region. A direct association was found between the HDI and the difference in mortality attributable to DM (Spearman's rho = 0.41; p = 0.04). CONCLUSIONS: Mortality attributable to DM increased over the study period. It was highest in coastal areas, intermediate in the rainforest region, and lowest in the highlands. A direct association was found between the HDI and the difference in mortality attributable to DM.