Political involvement is conceptualized as an orientation toward a political situation that distinguishes qualitative differences in message processing. This stands in contrast to traditional conceptions of political involvement, which describe deep-seated individual traits that differ little over time. Three categories of situational political involvement are proposed (active involvement, passive involvement, and uninvolved) and accompanying communication patters are examined with a 52-item Q sample during the 1994 California gubernatorial campaign. The actively involved category indicated different communication behaviors than the passively involved - and especially the uninvolved. The most important behaviour distinguishing the actively involved was interpersonal communication - talking about politics.