Following a work refusal at a plant manufacturing ice cream novelties in Ontario, we were asked to document cases of cumulative trauma disorders (CTDs) and carpal tunnel syndrome (CTS) in this workplace. There were 17 employees with possible hand and wrist problems identified from Workers Compensation Board (WCB) Forms, and from a list prepared at the time of the refusal. After obtaining consents, confirmations of the diagnoses of CTDs, CTS, and of surgical procedures for CTS were obtained from the physicians involved. The relative risk for these disorders among plant employees was estimated in two ways: 1) the rate of CTS operations between 1979 and 1990 was compared to that in the general population using Ontario Health Insurance Plan (OHIP) data on physicians' billings for these operations; and 2) the frequency of WCB first payment claims for tendinitis and CTS during 1987 to 1989 at the plant was compared to that among the entire labor force of Ontario. CTDs had been diagnosed in all 17 workers: 9 had had operations for CTS, but one had had this operation prior to working at the plant. Compared to the remaining 8 workers who had CTS operations, an estimated 0.08 CTS operations would be expected among the 150 employees on the plant's seniority lists between 1979 and 1990, if the estimated rates in the general population were present at the plant, giving a Standardized Morbidity Ratio of 10.0 (95% confidence interval [CI] 4.3-19.7; one-sided p = 2.1 x 10(-6)). There were 6 WCB claims for tendinitis and CTS among plant employees during 1987 through 1989. This frequency was about 68 times that in the entire Ontario labor force (95% CI 24.7-150). This investigation has shown that CTDs, and particularly CTS, documented by medical records, have occurred at least 10 times more frequently than expected at this plant. Use of health insurance billing data to estimate CTS operation rates represents a simple method for estimating the burden of illness at the individual plant level due to CTS (at least for that portion proceeding to surgery), using an objective outcome that can be confirmed from medical records.