Abstract Formulae for translating Hi-Vol data into estimates of inhalable particles ( IP ⩽ 15 μm) and fine particles ( FP ⩽ 2.5 μm) are developed, evaluated and applied. The equations re developed using simultaneous data from dichotomous samplers and Hi-Vol samplers at 75 locations nation-wide. The formulae are multivariate in the sense that they include the Hi-Vol parameters, TSP, SO 4 2− and Pb; the formulae are hybrid in the sense that the coefficients are partly physicochemical and partly statistical. Several sets of equations are presented with varying degrees of complexity. The Hi-Vol parameters are added in a stepwise fashion—TSP, then SO 4 2−, then Pb. Also, there are national aggregate equations (e.g. IP = 0.61 TSP or FP = 0.30 TSP) as well as equations disaggregated by site-type, region and region/season. Depending on the level of complexity, the predictive errors are as follows: 26–31% for individual daily values of IP, 13–16% for annual mean values of IP, 39–56% for individual daily values of FP and 16–30% for annual mean values of FP. A major application study using five years of Hi-Vol data at 226 California sites allows an investigation of the statistical, geographical and seasonal patterns of TSP, IP and FP throughout California. The most salient features of the application study involve the extremely high particulate concentrations in the Los Angeles area and the San Joaquin Valley.