After an ascending trend of cardiovascular mortality, which was observed up to the sixties, there was a declining tendency in western countries thereafter. The decrease in mortality rates in Poland has been noticed since 1991. There is an uncertainty whether this improvement in prognosis was only due to the improvement in health care level or was accomplished, to some extent, by population life style changes. To assess whether the decrease in cardiovascular mortality in Poland was accompanied by a simultaneous reduction in global cardiovascular risk profile. The study was based on screened random samples from the Warsaw population aged 35-64 years, who were examined in the years 1984 (2646 subjects), 1988 (1433 subjects), 1993 (1539 subjects) and 2001 (853 subjects). The group evaluated in 1984 was followed up for 10 years with all fatal events recorded (364 deaths were registered including 166 cardiovascular deaths). The model of 10 years probability of total and cardiovascular death was developed, which was based on 11 risk factors (age, smoking cigarettes, systolic blood pressure, pulse pressure, ratio of total cholesterol/HDL-cholesterol, triglyceride, symptoms of coronary heart disease or heart failure, death of mother before 65 or death of father before 55 years due to myocardial infarction or stroke, energy in daily food intake and percentage of energy derived from saturated fatty acids). This model was applied to assess the probability of deaths between years 1984 and 2001, utilising data from screenings. The probability of death (in %) decreased from 1984 to 2001 by 11% in men. However, it remained stable for women. On the other hand the probability of cardiovascular death decreased by 25% and 33%, respectively. The main role in global risk decrease was played by beneficial trends in mean systolic blood pressure (in both genders) and percentage of smokers in men. Modification of life style influencing risk factors profile decreased the probability of death in the Warsaw population, correlating with changes in mortality rates.