With whom and how frequently do people discuss political matters? A theoretical model is formulated and tested using information on respondents' core social networks from the 1987 General Social Survey. The frequency of political discussions depends upon factors of demand (personal attributes) and supply (opportunities within social contexts). Educational attainment and age positively affect political discussion frequency, although indirectly through intervening political orientation and social contexts variables. Demand is positively related to political interest, strong party attachment, liberal views, and newspaper readership. The findings underscore the importance of including close relatives and friends (strong ties) in studying social contexts of political behavior. The opportunity structure (supply of potential discussants) is enhanced by the presence of a spouse and other adult household members, by being in the labor force, and by having close personal relationships with spatially proximate core associates.The classic Columbia voting studies (Berelson, Lazarsfeld, and Mc-Phee 1954;Lazarsfeld, Berelson, and Gaudet 1948) revealed the role of primary-group contacts (family, co-workers, and friends) as providers of political information and influence. Despite the importance of personal networks, little is known about the factors that promote and restrict the flow of political information and influence (Sheingold 1973, pp. 714-15; Weatherford 1982, pp. 119-20).