Background. In earlier studies, serum total cholesterol was inversely associated with death due to respiratory diseases. Methods. Relationships of cholesterol with risk of respiratory diseases were examined between 1978 and 1993 in a health maintenance organization in Northern California (48,188 men, 55,276 women, about 60% white, age range 25-89). Incident and fatal endpoints were ascertained using hospital discharge diagnostic codes and underlying cause of death, respectively. Results. An inverse association of cholesterol level with pneumonia/influenza hospitalization was found among men and women except that, contrary to hypothesis, a positive association was seen in women under 55 years of age. Inconsistent inverse associations were found in men, but not women, between cholesterol level and hospitalization for chronic obstructive pulmonary disease (COPD). Hospitalization for asthma was unrelated to cholesterol level. An inverse association existed between cholesterol level and hospitalization for miscellaneous respiratory diseases among men, but not women. Exclusion of early hospitalizations tended to attenuate risk for cholesterol level below 4.14 mmol/l, but not the risk gradient across higher cholesterol levels. For deaths, cholesterol level was statistically significantly inversely associated only with COPD. Conclusions. Cholesterol had graded inverse associations with some respiratory diseases, but not others, tending to be stronger in men than in women, and in older than in younger participants. Associations varied for hospitalizations compared to deaths. More epidemiological and basic research is needed to elucidate the potential mechanisms underlying the observed associations.