In a case control study using a standardized protocol, 600 subjects were evaluated for essential tremor (ET). We demonstrated that ET was significantly more frequent in patients with Parkinson's disease (PD) (12/204, 5.9%) compared to diseased controls (2/206, 1%) and healthy controls (1/190, 0.5%). A regression analysis with ET as outcome and group (either PD or healthy controls or diseased controls) as independent variable (adjusting for age and sex) revealed that PD had higher odds of having ET than diseased controls (OR = 5.43, 95% CI = 1.16, 25.39, P < 0.001) and healthy controls (OR = 10.87, 95% CI = 1.39, 85.15, P < 0.001). The low frequency of ET in our controls was further confirmed in a follow-up study in a group of age and gender matched general medical patients who attended an outpatient clinic (0% frequency). Eight of 204 PD (3.9%) compared to none of diseased (0%) (P = 0.004) and healthy controls (0%) (P = 0.008) had a prior diagnosis of ET. The duration of ET symptoms in patients with PD was 25.1 +/- 19.6 (range 3-60) years. A multivariate analysis demonstrated that a lower dose of levodopa (OR = 0.993, 95%CI for OR = 0.988, 0.997, P < 0.001) and a higher age of onset of disease (OR = 1.108, 95%CI for OR = 1.035, 1.187, P < 0.001) were associated with increased odds of PD with ET, compared to patients with PD without ET. In our Asian population, patients with PD were 5 to 10 times more likely to have ET compared to diseased and healthy controls, suggesting that the association of ET and PD is unlikely to be ethnicity-specific.