In order to further examine a preliminary association between smoking and the subsequent development of Warthin's tumor, a case-control study was undertaken between the years 1980 and 1989. Consecutive patients with Warthin's tumor served as cases, and consecutive patients with pleomorphic adenomas served as controls. A positive history of smoking was defined as a patient with greater than a 10-pack year history. There were 28 cases and 69 controls (ratio: 1:2.5). The number of male to female patients with Warthin's tumor and pleomorphic adenoma was 22 to 6 (3.7:1) and 22 to 47 (1:2.1), respectively. The mean age at presentation with Warthin's tumor or pleomorphic adenoma was 60.7 years and 40.7 years, respectively. In those patients for whom a definite smoking history was known (93 patients), an odds ratio of 8.1 (p < 0.001) was calculated. In order to exclude sex as a possible confounder, the odds ratio in males only was 6.4 (p = 0.007). In order to exclude age as a possible confounder, the odds ratio in all patients over 45 years (48 patients) was 11.2 (p < 0.001). In summary, smokers have eight times the risk of developing a Warthin's tumor than nonsmokers. This case-control study suggests that smoking may be one of the etiologic factors associated with this benign salivary gland tumor.