OBJECTIVES: Syst-China is the ongoing placebo-controlled double-blind outcome trial in older (aged 60 years or more) Chinese patients with isolated systolic hypertension (systolic blood pressure 160-219 mmHg and diastolic blood pressure < 95 mmHg). This article is based on the data accumulated until 31 August 1992. Its purpose is to investigate the extent to which the variability in the clinic blood pressure readings at baseline could predict the blood pressure changes observed in the placebo arm of the trial. METHODS: From 2379 patients recruited into the trial, 728 [455 men and 273 women, aged 66.7+/-5.5 years (mean +/- SD)] were selected, because their blood pressure readings for the three run-in visits as well as 3, 6 and 12 months after random allocation were available. Overall and between-visit blood pressure variabilities at baseline were estimated from the two readings obtained with the subject seated during the first and second run-in visits. The baseline blood pressure used to calculate the blood pressure changes during follow-up was the average of the two readings during the third run-in visit. RESULTS: The blood pressure variability at baseline was larger for women than it was for men. For all of the subjects combined, the blood pressure had decreased by 4.1+/-14.4 mmHg (P < 0.001) systolic and 0.5+/-6.7 mmHg (P < 0.06) diastolic by the 3-month follow-up visit, by 8.5+/- 15.2 and 1.4+/-7.5 mmHg, respectively, after 6 months and by 10.3+/-15.7 and 1.9+/-7.9 mmHg, respectively, after 1 year (p < 0.001 for all). Stepwise multiple regression analysis showed that sex, age, alcohol intake and the blood pressure at baseline were significant determinants of the long-term (1 year) blood pressure changes. Aftger adjustment for the aforementioned covariates, the between-visit variability was a significant predictor of the changes in the diastolic blood pressure after 1 year of placebo treatment for the men (partial r+/-SEM -0.36+/-0.12, P < 0.01) and for all of the subjects (-0.19+/-0.09, P < 0.05). For men, the partial regression coefficient between the overall variability and the changes in the diastolic blood pressure also attained statistical significance (-0.39+/- 0.14, P < 0.01). CONCLUSION: For older Chinese patients with isolated systolic hypertension, in particular for men, a higher blood pressure variability at baseline was associated with a larger decrease in diastolic blood pressure during 1-year follow-up on placebo, explaining up to 2% of the variance of the observed changes. Similar associations were not observed for systolic blood pressure.