The importance of various factors in explaining geographical BP variations in the UK has been examined among 2,596 men and women living in nine British towns. Associations between BP and potential explanatory variables have been examined first for individual subjects (within-population), and second for towns (between-populations). The factors showing associations with BP that were consistent within-population and between-towns were BMI, urine sodium/potassium ratio (Na/K), alcohol and anxiety at examination for men, and BMI, Na/K and anxiety at examination for women. After adjustment for these factors and age, the male BP differences between the nine towns were reduced from 10.1 mmHg systolic in men (P = 0.001) to 7.1 mmHg (P = 0.04), and from 5.9 mmHg diastolic (P = < 0.0001) to 5.5 mmHg (P < 0.0001). For women, adjustment only marginally reduced the between-town systolic differences from 5.6 mmHg (P = 0.05) to 5.5 mmHg (P = 0.3) and increased the diastolic differences from 3.7 mmHg diastolic (P = 0.03) to 5.0 mmHg (P < 0.0001). It is concluded that BMI, alcohol consumption and the Na/K ratio play an important part in the pattern of male SBP variations in the UK. The factors associated with female geographical differences in BP were less clear, but BMI and Na/K ratio appear to be important. This study has been unable to identify factors associated with geographical differences in DBP, with the possible exception of BMI in men. However, as systolic pressure is recognised to be more strongly associated with cardiovascular outcome than diastolic pressure, it can be recommended that primary prevention strategies aimed at reducing the population mean level of BP should involve efforts to reduce overweight, alcohol consumption and the Na/K ratio.