A time-series study was conducted to ascertain the short-term effects of different-sized airborne particulate matter (PM) on daily respiratory and cardiovascular cause-specific mortality in winter and summer, among subjects aged over 75 years in Madrid. Poisson regression was used to analyse the time-series, in which the dependent variable was daily mortality due to different specific respiratory and circulatory causes, and the principal independent variables were daily mean PM10, PM2.5 and PM10-2.5 concentrations; other variables: other air pollutants (chemicals, biotic and acoustic), influenza, trend, seasonality and autocorrelation of the series. The results indicated an association between coarser PM fractions (PM10 and PM10-2.5) and respiratory-specific mortality on the one hand, and between PM2.5 and cardiovascular-specific mortality on the other. While the risk of mortality due to exposure to particulate matter was greater in summer than in winter, this difference was statistically significant solely for total organic-cause mortality.