The pattern of geographic blood pressure variations in Britain has raised the possibility that temperature or other climatic factors may be of importance. Data from two population studies have been examined: the British Regional Heart Study (BRHS), which involved 7735 mean aged 40-59 years, and the Nine Towns Study (NTS), concerning blood pressure among 2596 men and women aged 25-59 years. In the BRHS, significant negative associations were found between daily maximum outdoor temperature and systolic blood pressure (-0.38 mmHg/degrees C; P less than 0.001) and diastolic blood pressure (-0.18 mmHg/degrees C; P less than 0.001). There were similar, although non-significant, associations in the NTS. No significant associations were found between blood pressure and room temperature in either study after taking account of town blood pressure differences, nor between blood pressure and outdoor humidity in the NTS. In the NTS, skin temperature was negatively associated with blood pressure after the adjustment for body mass index, significantly so for male diastolic (-0.62 mmHg/degrees C; P less than 0.05). The BRHS estimates suggest that, in Britain, geographic differences in outdoor temperature may contribute no more than 2 mmHg systolic and 1 mmHg diastolic to regional blood pressure variations. This represents a relatively small proportion of the town differences in blood pressure observed in both the BRHS and NTS. Furthermore, international comparisons suggest that environmental temperature is not an important determinant of population blood pressure levels.