The results of the dietary aflatoxin-liver cancer study carried out in the Murang'a district of Kenya have been reassessed in relation to disease incidence rates based on a total of seven years of cancer registration and related to the Population Census carried out during the course of the initial study. These newly derived data have been combined with the results of a second similar dietary aflatoxin-liver cancer study which was later carried out in Swaziland. Separate treatment of the male and female data has been considered necessary due to the variation of the sex ratio of the disease incidence in the two areas. The combined results of the two studies show a high degree of positive correlation between the calculated ingestion levels of aflatoxin, expressed as ng/kg bodyweight/day (x) and the adult incidence rates of hepatocellular carcinoma expressed as cases/10(5) adults/year (y) within the two study populations and this is true for both males and females. Based upon an assumed 2 kg/day intake of wet diet and a mean bodyweight of 70 kg, the calculated relationship for adult females is: y = 4.14 log10 x--0.80 (0.05 greater than P4 greater than 0.02, tor = +/- 2.90). With the added assumption of a daily intake of native beer of two liters/day the regression equation for adult males is: y = 21.96 log10 x--11.17 (P5 less than 0.001, tor = +/- 3.42). The regression data are found to be essentially compatible with comparable data recorded by independent workers in Thailand and Mozambique; the latter being the region where the highest rates of liver cancer and of aflatoxin ingestion levels have so far been recorded. A highly significant regression line has also been calculated using crude disease incidence data and aflatoxin exposure levels from all available studies: y = 7.60 log10 x--3.60 (P8 less than 0.001, tor = +/- 3.10).