Five thousand five hundred and forty six cases of all lung cancer patients who died during 1980-1988 in Guangzhou, China were investigated retrospectively with a standardized 31-item questionnaire administered to their next of kin. The data of 1093 lung cancer patients(20%, 1093/5546) who had a histological record was analyzed to investigate the changes in histological types and the possible etiologic causes. The difference between the lung cancer deaths with and without histological record is not significant in age, location (peripheral or central) of tumour and status of occupation (P > 0.05), but the difference in sex is significant (P < 0.01). We analyzed the data of 1093 cases by sex and by 3-year period. There had been a shift in the histology pattern with an increase in the percentage of adenocarcinoma (P = 0.0011) and a decrease in percentage of squamous cell carcinoma (P = 0.0027) in males, inversely, there has been an absolute and a relative decrease of percentage in adenocarcinoma in females (P = 0.0028). The percentage of smokers, age of starting to smoke and type of tobacco smoking were nearly constant in both sexes during the studied periods. The pack-years of smoking decreased in males (P = 0.0396), and seemed increase in females (P = 0.1576, no significance). The analysis of occupation and dietary habits among 5546 cases were performed. The proportion of housewives decreased with time (P < 0.001) while the percentage of chemists in females increased (P < 0.001) with time. The dietary habits are changing with an increase in roast food intake for males (P = 0.0055) and in vegetable intake for males (P < 0.0001), females (P < 0.0001), and for patients with lung squamous cell carcinoma and adenocarcinoma in both sexes (P < 0.001). Perhaps the changes in pack-years of smoking may be responsible for the percentage change of lung cancer histological types observed in our study. The role that changes in dietary habits and status of occupations play in the changing trends of lung cancer histological types in our study need further study.