A composite receptor model for PM-10 apportionment was evaluated to determine the stability of its solutions and to devise cost-effective measurement strategies. Ambient aerosol samples used in the evaluation were obtained with dichotomous samplers at three sites in the vicinity of Philadelphia, PA and were analyzed by x-ray fluorescence (XRF), instrumental neutron activation (INAA), scanning electron microscopy (SEM) and ion chromatography. The composite receptor model consisted of a wind trajectory method, chemical mass balances (CMB), and multiple linear regression (MLR). In resolving PM-10 into 11 components, SEM was essential for coal-fly ash and botanical matter, XRF was essential for a S-component and vehicle emissions, and INAA was essential for fluidized catalytic crackers at refineries. The remaining components were determined well by either INAA or XRF. Ten to twenty samples were sufficient to determine average source contributions by CMB. Multiple linear regression required 30 or more samples to determine the S-component and at least 50 samples for vehicle emissions.