Contradictory findings exist on the political party preferences of intergenerationally occupationally mobile men. The argument is recast in terms of additive versus interactive models of origin and destination occupational group effects. Data from American presidential year surveys in the 1960s reveal that an additive model satisfactorily explains variance in party identification among those who are occupational stable as well as among those who are mobile; mobility per se has no effect. Additional analyses suggest that a model of selective recruitment for mobility provides a better account of the observed pattern than does a model of resocialization: the partisan-ship of those who are mobile is intermediate between those who are stable in their class of origin and those stable in their class of destination.