Determinants of change in blood pressure between two examinations separated by a 6-year period were investigated in 340 working men aged 20-45 at first examination and not on antihypertensive medication at either examination. A London School of Hygiene sphygmomanometer was used for all measurements of blood pressure. Relationships between change in blood pressure (residuals of follow-up blood pressure regressed on baseline blood pressure) and changes in weight, physical activity, cigarette smoking, diet, alcohol intake and a range of psychological characteristics were examined using multiple regression. Systolic and diastolic changes in blood pressure were both positively related to baseline age and weight. Regression equations indicated independent effects of change in weight, with similar increases of 0.5 (s.e. 0.1) mmHg in systolic and diastolic blood pressure for each kg of weight gain. Decreased coffee consumption and increased participation in sports were negatively associated with systolic changes in blood pressure. Men who had stopped smoking cigarettes were observed to have higher than average increases in both systolic and diastolic blood pressure, although these relationships were not independent of other variables. The difference in ambient temperature between baseline and follow-up measurement days was an important determinant of change. Changes in mean systolic and mean diastolic blood pressure in 76 manual workers (+10.1/+6.4 mmHg) compared unfavourably with changes in professional men (-2.8/+2.7 mmHg).(ABSTRACT TRUNCATED AT 250 WORDS)