Mortality from smoking-related diseases is rising steeply among women. Although most studies recognize that smoking habits differ between men and women, this knowledge is seldom applied when studying gender differences. Based on two large population studies comprising almost 16,000 study subjects, we illustrate the importance of adjusting for smoking in detail when comparing hospitalization and mortality between men and women. Adjusting for smoking as a dichotomous variable, age-adjusted mortality risk during 16 years follow-up was 80% higher for men than women. Adjusting for smoking in detail significantly reduced men's excess risk to 50%. The risk of hospitalization for chronic obstructive pulmonary disease was 50% higher for men when adjusting for smoking as a dichotomous variable, whereas with full adjustment the risk was reversed to a significantly increased risk for women of 20%. We conclude, that when comparing smoking-related morbidity in men and women, smoking habits must be adjusted for in more detail than is usually the case.