Much recent work has focused on Americans’ positive and negative feelings toward their own and opposing political parties. However, there is neither a consensus on how to model such partisan affect, nor a detailed understanding of its consequences for political participation. This work addressed these two gaps by first, empirically examining how many dimensions best characterize American partisan affect. Study 1A used contemporary, categorical approaches to factor analyses across an extensive set of partisan affect items from the Pew American Trends Panel to test two competing hypotheses: that (1A-1) partisan affect is one-dimensional, or that (1A-2) partisan affect is two-dimensional. Results suggested support for Hypothesis 1A-2; two dimensions of partisan affect covered inparty affinity and outparty animosity. Second, Study 1B investigated the predictive validity of different aspects of partisan affect in terms of discrete forms of political participation. Study 1B had three competing hypotheses implied by prior partisan affect literature: that (1B-1) outparty animosity (but not inparty affinity) would predict most forms of behavior, that (1B-2) outparty animosity would predict higher-cost forms of behavior, and inparty affinity would predict lower-cost forms of behavior, or that (1B-3) the combination/interaction of outparty animosity and inparty affinity would predict most forms of behavior. Results of logistic regressions suggested partial support for Hypothesis 1B-1 and direct support for Hypothesis 1B-2. Outparty animosity predicted more medium-cost forms of political participation, whereas inparty affinity predicted lower-cost forms of political participation. Implications are discussed for theory, the measurement of partisan affect, and the prediction of political participation.