PURPOSE: In urban Shanghai, the largest industrial and commercial city in China, the age-adjusted (world standard) incidence rates for colorectal cancer increased from 14.5 to 23.3 per 10(5) men and from 12.1 to 20.3 per 10(5) women between 1972 and 1996. This change was even more pronounced for colon cancer, whose incidence rates doubled from 5.95 to 13.7 per 10(5) men and from 5.77 to 12.5 per 10(5) women. The reasons for the rapid increases in cancer rates are not fully understood, but may involve dietary exposures that have changed substantially over the past two decades.METHODS: We calculated Pearson correlation coefficients (r) between colorectal cancer rates and the dietary factors of grain, vegetable oil, pork, poultry and vegetable consumption over the period of 1972 through 1996 in urban Shanghai.RESULTS: Statistically significant positive associations were observed between colon cancer rates and per capita consumption of vegetable oil (r = 0.91 for men, r = 0.94 for women), poultry (r = 0.90 for men, r = 0.90 for women), and pork (r = 0.78 for men, r = 0.81 for women). The correlation coefficients were not statistically significant between colon cancer and per capita consumption of grain (r = 0.38 for men, r = 0.37 for women) or vegetables (r = 0.16 for men, r = 0.14 for women). Similar weaker associations were observed between rectal cancer rates and vegetable oil, pork and poultry consumption.CONCLUSIONS: The findings in our study suggest that increases in dietary fat, poultry and pork intake may play a role in the rising colorectal cancer rates in Shanghai.