The evidence for a hypothesis that eating salted fish is associated with nasopharyngeal carcinoma (NPC) is reviewed. The hypothesis was tested among Malaysian Chinese using a matched case-control design. The kinds of salted fish and patterns of use were also investigated in a control group comprising 100 Chinese, 50 Malay and 50 Indian households. During 1980, in Selangor, Malaysia, interviews with 100 Chinese cases of NPC and 100 non-disease controls indicated that salted fish consumption during childhood was a significant risk (relative risk = 3.0, P = 0.04), with an elevated risk for daily as opposed to less frequent consumption. Salted fish consumption during adolescence was a less significant risk, and current consumption not at all. There were 19 kinds of fishes reported as being eaten as salted fish by the 200 control households. There were marked differences between ethnic groups in preference for different kinds: Chinese preferred red snapper (74% of households), Malay jewfish (54%) and Indian red snapper (28%). Salted fish was hardly ever eaten daily by any household; weekly was a moderate frequency in all ethnic groups; less than weekly most common. There were no statistically significant differences between Chinese NPC case and non-disease control participants in kind of salted fish eaten. Results were the same when the data were analyzed by sex, subethnic group and income.