Short term effects of air pollutant levels on daily mortality were studied, using time series approach, in Milan from the year 1980 to 1989. The Poisson regression with autocorrelated residuals was applied. This paper reports preliminary results of the analyses on the association between total daily number of deaths and the 24h mean concentrations, on the same day, for two air pollutants: sulphur dioxide and total suspended particulate. A positive and curvilinear relation (with a logarithmic shape) has been found between total daily mortality and concentration levels for each pollutant. This result confirm a steeper increase in mortality at low concentrations, lower than the international standards in use at the moment.