Estimates of an individual's ozone exposure were generated for a panel of 51 asthmatic study subjects using ambient ozone concentrations, the relationships observed between the ambient concentrations and the concentrations found in each of two broad classes of environments and knowledge of the individual's movement through these locations. These estimates are compared to actual ozone concentrations observed during personal monitoring of 30 of the individuals. Earlier comparisons disclosed that the estimated hourly means were in much better agreement with the observed hourly means than were the ambient hourly values. The hourly maxima observed in the personal monitoring for each of three environments are compared to the exposure estimates. The ability of the exposure model to estimate average hourly exposures for individuals has been previously tested. In the present work, the ability of the exposure model to estimate peak exposures within an hour is evaluated, using measurements of ozone concentration obtained at 5 minute intervals during the personal monitoring program as a standard. One-hour exposure estimates based on fixed site measurements are evaluated as well. Sources of error within the data set are identified but appear to not seriously affect the results. The exposure model produces one-hour estimates which are 7-12 ppb below the observed personal exposures.