A study tested the knowledge gap hypothesis, a promising framework for research in the field of mass communication devised by P. Tichenor, G. Donohue, and C. Olien in 1970. The study investigated the relative contribution of two types of factors--structural and motivational--in predicting the degree to which persons will attend to health messages, and differentials in levels of knowledge about cancer. Secondary analysis was conducted on data collected in 1983 from a probability sample of Wisconsin households in which professional interviewers had completed 1,050 interviews at a response rate of 65%. Results indicated that motivation and focused exposure to health information are significant predictors of knowledge about cancer, and that the frequency of attending to media messages about health is not structurally determined (predicted by socioeconomic status, education and income), but instead, the result of concern about or interest in a person's own health. Results also revealed important differences between erroneous and correct knowledge, corroborating previous findings on knowledge typologies. These findings may validate information campaigns and the notion that people can be motivated to seek health information independent of their social location; they may also validate the wide variety of ongoing cancer information campaigns within that context. (Three figures and 4 tables of data are included; 26 references are attached.) (PRA)