Objective To prospectively explore the association between tea drinking and incidence of stroke of adults of Zhejiang province.Methods After excluding participants with heart disease,stroke,cancer and diabetes at baseline study,53 916 participants aged 30-79 years in the China Kadoorie Biobank (CKB) study from Tongxiang were included for final analysis.Cox regression model was used to estimate the hazard ratio (HR) for the association of tea drinking with incident stroke.Results The main type of drinking tea was black tea (79.78％),followed by green tea (20.08％).Of the 53 916 participants,the proportion of participants who drank tea at least once per week was 31.27％.The corresponding proportions for men and women were 60.24％ and 10.30％,respectively.Among 391 512 person-years of the follow-up program (median 7.26 years),a total of 1 487 men and 1 769 women were diagnosed with stroke.After adjusting for socio-demographic status,lifestyle,BMI,waist circumference,and systolic blood pressure,HR for incident stroke decreased with the increase of daily average tea consumption amount (P=0.000 6).Compared with participants who did not drink tea weekly,the HRs for incident stroke in those consuming tea 0.1-,3.0-and ≥5.0 g/d were 0.93 (95％ CI:0.85-1.00),0.88 (95％ CI:0.77-0.99) and 0.79 (95％ CI:0.69-0.89),respectively.The HRs for incident stroke in smokers and non-smokers who consumed tea ≥5.0 g/d were 0.71 (95％CI:0.59-0.86) and 0.97 (95％CI:0.77-1.21),respectively,compared with current smokers and non-smokers who did not drink tea weekly (P=0.040 0).The corresponding HRs for alcohol drinkers and non-drinkers were 0.96 (95％CI:0.76-1.22) and 0.70 (95％CI:0.58-0.84),respectively (P=0.040 0).The corresponding HRs for central obese persons and non-central obese persons were 0.60 (95％ C1:0.44-0.81) and 0.86 (95％ CI:0.73-1.01),respectively (P=0.040 0).Conclusion Tea drinking had an effect on reducing the possibility of incident stroke,especially among those who were current smokers,non-alcohol drinkers and central obese.