Participants filled 42 orders on a task designed to simulate components of filling prescriptions. Task factors included objective workload of 70- versus 80-min. to complete the task and perceptions of workload dimensions using the NASA Task Load Index. The proportion and pattern of data-entry, counting, and product-selection errors were compatible with those found in pharmacy field-sites. Significant other relationship stress, field-dependence, and an 80-min. workpace predicted data-entry errors. Mistakes in product selection were associated with low GPA, high social stress, the NASA Task Load Index dimension of less concern with performing well, and a 70-min. workpace. Relationship of data to corresponding information in the pharmacy literature and to assumptions of a cognitive-systems performance model was discussed.