The prevalence of chest symptoms in blue collar workers not known to be exposed to hazardous occupational respiratory conditions was studied. A population of 1372 workers from 14 facilities in North Carolina was studied. All were from environments as free as possible of respiratory hazards. A standard respiratory symptom questionnaire was administered. Linear logistic prediction equations were developed for nine symptoms categories. For each smoking status, current or previous smokers, or those who never smoked, equations were fitted employing race, sex, age, height, weight, and education as dependent variables. Symptom prevalences ranked by age group for smokers, exsmokers, and those who did not smoke yielded prevalences from 1 to 8 percent for moderate dyspnea among males depending on smoking status and race; 3 to 13 percent for periods of increased cough or phlegm, depending on smoking status and race; and 0.0 to 17 percent for wheezing, depending or smoking status, race, and weight. The authors conclude that descriptive statistics can be used as a guide for selecting risk factors to adjust for symptom prevalence in population studies.