Low intake of foods rich in vitamin C is associated with an increased risk of gastric cancer, and geographic variation in average vitamin C intake, therefore, could explain some of the wide international variation in gastric cancer rates. This multicentre study investigated the relationships between plasma levels of vitamin C, as an indicator of vitamin C intake, and gastric cancer rates, markers of gastritis and other socio-demographic variables. Fasting plasma samples from about 1,400 individuals from 9 centres in 7 countries worldwide were assayed for total vitamin C using a fluorometric assay. There was no association between average plasma vitamin C levels and either gastric cancer mortality or incidence rates in the populations studied. Therefore, variation in fasting plasma vitamin C levels, as an indicator of consumption of vitamin C, does not appear to explain any of the wide geographic variation in gastric cancer rates. Furthermore, there was no association between plasma vitamin C levels and Helicobacter pylori infection, low serum levels of pepsinogen A (as a marker of severe chronic atrophic gastritis) or the presence of DNA adducts in blood leukocyte DNA. Multivariate models showed that fasting plasma vitamin C levels were associated positively with female sex, higher levels of education, never having smoked and increasing height and negatively with number of cigarettes smoked per day and increasing weight. This suggests not only that gender and tobacco smoking, in particular, are important predictors of plasma vitamin C levels but also that their effects are consistent throughout the developed world.