Risk factors for Japanese cedar pollinosis including past or family history of allergic diseases, smoking and passive smoking, dwelling conditions, and life events were analyzed by a case control method. Patients with Japanese cedar pollinosis (22 males and 67 females) were matched with a corresponding number of patients without potential symptoms of pollinosis according to sex and age (+/- 5 years). The mean age was 39 years in both groups. The odds ratio (OR) was calculated by McNemar's method and the conditional logistic regression model. The design and methodology in this study were somewhat inadequate so that the validity of the results is limited. The most important problem was no-matching according to exposure to pollen. Significantly high OR for past history of allergic disease (8.80, 95% confidence interval (CI); 3.49-22.2), atopic sermatitis (9.00, 95% CI; 1.14-71.0), and a sibling history of allergic disease (3.25, 95% CI; 1.06-9.97) were consistent with former genetical studies. ORs were lower than unity for current smokers (0.36, 95% CI; 0.11-1.13) and those smoking 10 cigarettes/day or more (0.20, 95% CI; 0.04-0.91) relative to nonsmokers. The OR for passive smoking from 7-15 years of age as a result of the father's smoking habit (0.38, 95% CI; 0.17-0.86) was also significantly low. Smoking was suggested to increase the level of total and antigen-specific IgE in serum by former studies, so that sensitization and symptoms should be studied separately. The high OR of residents in a business or light industrial area (5.00, 95% CI; 1.45-17.3) suggested an association with air pollution.(ABSTRACT TRUNCATED AT 250 WORDS)