The purpose of this study was to validate the proposed NIOSH definition for occupational asthma as useful for the identification of cases to be reported under NIOSH's Sentinel Health Notification System for Occupational Risks (SENSOR) program. The specific objectives were to assess the availability of information required for use of the case definition in occupational health clinic charts and to assess the frequencies with which the clinic primary physicians, an independent occupational medicine physician reviewer, and the surveillance case definition agreed with each other on the diagnosis of occupational asthma. The definition calls for the presence of symptoms suggestive of asthma (wheezing, dyspnea, cough, chest tightness which are variable or intermittent); documentation of significant reversible or variable airways obstruction; and evidence of an association between the pattern of airways obstruction and some workplace exposure. Charts were reviewed for 101 patients, of whom 39 were diagnosed as having probable or definite occupational asthma by the independent reviewer. Of the 63 patients with adequate information recorded in their charts to apply the proposed NIOSH surveillance case definition, 21 met the requirements of the case definition. In 84% of the 101 charts, the presence or absence of workplace exposure to a known agent was recorded. Evidence of work related change in pulmonary function were recorded on 17 of the 101 charts. Significant work related changes were documented in 11 (65%) of these 17. The authors concluded that there was variable and incomplete recording of data by the examining physicians with regard to the information required by the NIOSH proposed standard definition for occupational asthma. This suggests that the case definition as presented would be insensitive as a surveillance tool.