The mass media's role in informing the American public is critical to public support for government policies. The media are said to set the national agenda. This view is based on the assumption of selective coverage they give to news items. Media coverage also influences the salience the public attaches to issues. However, media agenda effect has been challenged by Lang and Lang (1983). These scholars, in their media agenda-building theory, argued that the success of media effect on national agenda is dependent on group support. In order to test this theory, time-related data on South Africa crises, media coverage"of South Africa, American public reactions, congressional, and presidential apartheid-related activities, between 1976 and 1988, were analyzed. Congressional anti-apartheid policies were the dependent and others, the independent variables. The theory made analysis of the data amenable to the additive adopted to test for the significance of the interactive variables, indicated that these variables were negatively related to congressional anti-apartheid policies. The additive model was subsequently analyzed. The time series multiple regression analysis was used in analyzing the relationships. Given autocorrelation and multicollinearity problems associated with time series analysis, the Arima (p, d, q) model was used to model the relationships. This model was used to indicate support, or nonsupport, for the time series regression analysis. The result of the additive model indicated that South African political crises were negatively related to congressional anti-apartheid actions. It also showed that the relationship between the American public reactions and congressional anti-apartheid policies was greater in comparison to all other independent variables. The presidential actions taken against South Africa were negatively related to Congress' anti-apartheid actions. Television had the greatest relationship with congressional anti-apartheid actions compared to newspapers and magazines.