It is established that personally monitored particulate matter (PM) is poorly correlated with ambient concentrations of PM. This observation is inconsistent with epidemiology studies relating ambient PM to mortality. This paper reviews in depth the relationships between personal and ambient PM concentrations from the literature, discusses possible reasons for low correlations between them, and the implications for interpretation of health effect studies relating mortality to ambient PM concentrations. This investigation shows that although ambient PM provides a poor predictor of an individual's personal PM, it explains much of the variability for a community's mean personal PM.