We present an epidemiologic approach for the study of the irritant effects of sodium borate exposure that can be used to study acute reversible health effects more generally. Acute irritant responses, such as symptoms of nasal irritation or cough, are subjective. They may, however, be accompanied by transient airway obstruction, which can be objectively measured by a change in peak expiratory flow (PEF) or forced expiratory volume in one second (FEV1). Because these physiologic responses are reversible and follow short-term peaks of exposure, they cannot be studied effectively using conventional epidemiologic methods of design and analysis. The proposed study design is a short-term prospective cohort study with frequently repeated measures of health effects and simultaneous measures of exposure. Because each subject passes through periods of high and low short-term exposure, the study has the form of a crossover design. This design permits each acute response to be linked to the corresponding short-term average exposure, providing the data necessary for dose-response analyses within subject. We present the results of the sodium borate study to illustrate the analysis of such data.