This study describes the geographical variations in the incidence of, and mortality due to, the most frequent cancers in Quebec. They are breast, colon and tracheal, bronchial, and lung cancers in women and prostate, colon and trachea, bronchial and lung cancers in men. Data for the periods 1984 to 1986 and 1989 to 1991 for the whole of the Quebec Community Health Department are reviewed. A direct method of standardization, with the world population as the reference population, was used to allow comparison of geographical units. The ranking correlation method was used to test correlations for the various cancer sites. There were significant coefficients between the two periods only for lung cancer for both sexes, for both incidence and mortality. The spatial autocorrelation method was used to determine the presence or absence of a geographical configuration from the estimated rates. Moran I values were positive and significant for the two periods for lung cancer associated mortality and for the incidence of breast cancer. This implicates local factors (probably associated with lifestyle). The results for the incidence of lung cancer among men and women were similar for the period 1989 to 1991. The advantages and limitations of the methods used and also the data available are considered.