Methods were developed for estimating fine particulates less than 2.5 microns in aerodynamic diameter (PM2.5) from airport visibility data which detected seasonality and allowed for possible site- and season-specific regressions of PM2.5 on visibility. The methods were applied to nine airports in California in order to produce estimates of PM2.5 for the years 1966-1986 based on 1,767 paired PM2.5/visibility data points where PM2.5 had been measured at a monitoring station near the airport. General F-tests indicated that site- and season-specific regression equations resulted in a statistically significant reduction in residual error. The split halves correlation for estimating PM2.5 from visibility over all areas was 0.82. The methods were used to estimate long-term concentrations of ambient PM2.5 for an epidemiological cohort of 1,868 individuals.