The adverse effects of extreme temperatures on mortality have been reported in areas with different climates. However, few studies have been conducted in countries in the tropical zone around the equator where the effects of extreme temperatures might differ. We aimed to estimate the association between ambient air temperature and natural-cause mortality and assess the impact of extreme temperatures on attributable mortality at a national and subnational level in Colombia during 2010-2019. We obtained data for daily temperature and mortality from 32 departments in Colombia and conducted a two-stage analysis. In the first stage we fitted a time-series Poisson model for each department and estimated the mortality-temperature association by using distributed-lag-nonlinear models with 7 up to 28 days of lag. In the second stage, we pooled these estimates using a multivariate meta-analytic model including temperature variables and a poverty index. We calculated attributable deaths and fractions due to non-optimal temperatures and due to cold and heat based on daily values below or above the department-specific optimal temperatures with lowest mortality. We analyzed 2,276,318 natural deaths and found marked differences in the exposure-response curve of mortality-temperature, where most departments showed heat effects but no cold effects for the 7-days lag. This lag-response curve for heat showed that the risk of death is higher during the same day (lag 0) of extreme temperatures and decrease after the third day. The country attributable fraction due to non-optimal temperature was higher for heat (1.68, 95% CI 0.91 – 2.38) during the 0-7 days lag, but higher for cold temperatures (4.72, 95% CI 1.53 – 7.40) during the 0-28 days lag. In conclusion, most of the acute temperature-related mortality in Colombia was due to heat and some departments showed higher risks. These results should inform planning adaptation strategies for climate change differentiated at subnational level.