The blood pressures of 252 men and 250 women, living in both urban and rural Zimbabwe, were measured on three separate occasions. Food and alcohol intakes were determined using a 3-day weighed diet survey checked by means of a detailed interview. Anthropometric data were also collected. No relationship was found between mean blood pressures and the alcohol intake for any socioeconomic group. Using only the first of the three blood pressure measurements, a correlation between systolic pressure and alcohol intake was found for white males (r = 0.234, P less than 0.05), and just missed statistical significance (r = 0.156, P = 0.065) for black middle class males. No relationship was found between blood pressure and alcohol consumption for black working class males, or for females. Epidemiological evidence suggests that alcohol consumption is associated with increased blood pressure. However, a convincing physiological mechanism is lacking. It is suggested that psychological factors may be partly responsible for this relationship.