BACKGROUND: The large differences in cardiovascular disease rates between Eastern and Western Europe have largely developed over the last few decades, and are only partly explained by classical risk factors. This study was set up to identify other potential determinants of these differences. METHODS: This was an ecological study comparing random samples of men aged 45-64 years selected from three cities representing populations with different rates of cardiovascular mortality: Pardubice (Czech Republic), Augsburg (Bavaria, Germany), and Jerusalem (Israel). In total, 191 (response rate 70\%), 153 (70\%) and 162 (62\%) men, respectively, participated. All centres followed the same study protocol. Lifestyle, anthropometry and biochemical risk factors were assessed by identical questionnaires, standardized medical examination, and central analyses of fasting blood samples. RESULTS: The mortality rates in the study populations, as well as the prevalence of coronary heart disease in study samples, were highest in Czech, intermediate in Bavarian and low in Israeli men. This pattern was replicated across the three samples by mean blood pressure (P \textless 0.001), cigarette smoking (not significant), triglycerides (P \textless 0.05), fibrinogen or D-dimer levels (P \textless 0.05). On the other hand, the prevalence of diabetes and obesity were similar; total and high density lipoprotein (HDL)-cholesterol, apolipoprotein B, lipoprotein (Lp(a)) and glucose did not differ between Czech and Bavarian men; and Czechs had particularly low levels of serum insulin and factor VIIc. Israelis had low fasting glucose and total cholesterol, as well as HDL-cholesterol levels and a high Lp(a) (each P \textless 0.001) compared with the two other samples. Striking differences were found for plasma homocysteine (10.5 in Czechs versus 8.9 mumol/l in Bavarians, P \textless 0.001) and for alpha-carotene (geometric mean in Czechs 16, Bavarians 21 and Israelis 30 micrograms/l), beta-carotene (60, 110 and 102 micrograms/l), and lycopene (84, 177 and 223 micrograms/l), respectively; all P-values \textless 0.001). Adjustment for obesity or smoking did not change these estimates. There were no differences in the levels of tocopherol and retinol. CONCLUSIONS: Czech men had high levels of blood pressure, triglycerides, fibrinogen and D-dimer but many other traditional risk factors, as well as indicators of metabolic disorders and vitamins A and E, did not differ between the study samples. The low levels of carotenoids and high concentrations of homocysteine in Czech men seem to reflect their low dietary intakes of fruit and vegetables. The results provide indirect support for the importance of dietary factors in the East-West morbidity and mortality divide.