In 386 men aged 40-49 years the number of platelets was related to differences in smoking habits. "Platelet adhesiveness" (retention) was estimated by a glass bead filter method in 376 of these men. The estimation of "adhesiveness" was performed in native blood without anticoagulants at least 12 hr after the last cigarette smoked. A small but statistically highly significant increase in platelet count, in number of "adhesive platelets" and percentage of "adhesive platelets" was found in smokers as compared with non-smokers, the highest values being found in the heaviest smokers and vice versa. Such smoking-related changes in platelet count and reactiveness might unfavourably influence the tendency towards coronary thrombosis, and might in part explain the deleterious effects of smoking on coronary heart disease morbidity and mortality.