Noise exposure has been associated with increased catecholamine production and blood pressure elevation in laboratory studies and in human volunteers. Epidemiologic studies have given conflicting results. In order to determine whether noise-induced hearing loss predicts a rise in blood pressure, we reviewed occupational medicine records in an occupational health center serving three companies where noise exposure is commonly found. Height, weight, blood pressure, and screening audiometry are obtained as part of routine occupational health screening, and the results of the screening visit are abstracted from written clinical records. The results of pure tone screening audiometry are reported in nonstandardized fashion (Normal, WNL, NAD, for example, for normal). We reviewed records from 1990 and 1991 inclusive. One investigator, blind to blood pressure status, assigned each record to "no hearing loss," "not codable," or "hearing loss assumed to be due to noise" on the basis of the written audiometry report. Hearing loss due to causes other than noise was considered not codable. No attempt was made to quantify severity of hearing loss. Two hundred and sixteen charts were excluded as "not codable," 1,535 were classified as having no hearing loss, and 610 had some degree of hearing loss, most probably due to noise exposure. To adjust for confounding covariates, multiple regression analysis was used and indicated that hearing status improves the regression model for predicting diastolic blood pressure (p = 0.04), following age, nationality, body mass index (BMI), and month of testing, although the effect is small. Stratification by age and BMI revealed increased diastolic pressure in the group with hearing loss under age 45, regardless of obesity.(ABSTRACT TRUNCATED AT 250 WORDS)