This article presents task-based exposure data measured on multiple workers within occupational categories and discusses (1) the importance of defining the basic unit of analysis and choosing an analytical plan consistent with it; (2) analytical approaches for estimating the task means and variances, including procedures for weighting the observations by measurement times; and (3) methods for estimating the mean and the variance of the mean for an overall occupational category, including methods for incorporating variability due to varying task proportions and dependence of worker exposures over tasks. The goal is to provide analytical techniques for summarizing task-level exposure data to the occupational group level. A simulation is used to illustrate important principles and show how the proposed methods perform better than other standard approaches. When the current survey data are used to estimate task proportions and worker dependency over tasks, a proposed jackknife method for computing the variance of the overall occupational category mean performs well. A proposed method for incorporating the variability in task proportions when their source is external to the current survey is also evaluated. When the variability is properly incorporated in the computations, stratification by task does not lead to any reduction in estimated variance of the overall occupational category mean. Weighting the observations by the measurement times provided no reduction in the variance estimates of the means.