Epidemiologic studies of irritants are difficult to perform using standard epidemiologic methods for several reasons, including the reversible nature of the health outcomes, the selection of sensitive individuals from the study population, and the wide heterogeneity in normal responses to irritants. This study examined the feasibility of using repeated measurements of peak expiratory flow (PEF) and reported symptoms to study respiratory irritants and their effects in students exposed to formaldehyde during a clinical anatomy laboratory course. We studied 24 physical therapy students dissecting cadavers for 3 h per week over a 10-wk period. Formaldehyde exposures in the breathing zone ranged from 0.49 to 0.93 ppm (geometric mean +/- geometric SD, 0.73 +/- 1.22). Irritant symptoms increased strongly over the course of the average laboratory period, but this effect was stronger at the beginning than at the end of the semester. PEF measured before each laboratory session declined over the semester by an average of about 10 L/min (2% of baseline), a trend that was statistically significant in random-effects regression models. After 14 wk away from the laboratory, the group's mean baseline PEF had returned to its preexposure level. Mean PEF also declined over each laboratory period, although this effect was attenuated over the course of the semester. Other important predictors of cross-laboratory PEF decrements were asthma and reporting throat irritation during the laboratory. It appears that mild irritant effects can be detected in naive subjects using a repeated monitoring design and relatively simple instrumentation.