We use longitudinal data incorporating three generations of Americans to reevaluate the character and consequences of political socialization within the family. Findings about parental influence based on youth coming of age in the 1990s strongly parallel those based on youth socialized in the 1960s. As expected on the basis of social learning theory, children are more likely to adopt their parents' political orientations if the family is highly politicized and if the parents provide consistent cues over time. The direct transmission model is robust, as it withstands an extensive set of controls. Early acquisition of parental characteristics influences the subsequent nature of adult political development.W riting 40 years ago, Jennings and Niemi (1968) questioned the conventional wisdom about the role of parents in shaping the political character of their children. Working from the perspective of social learning theory and drawing on data collected independently from adolescents and their parents, they demonstrated high variability in the political similarity between parents and their children. Especially when judged against the expectations laid down by reliance on retrospective accounts of parental attributes, the results appeared to downgrade the direct transmission model, wherein parental attributes were passed on, wittingly or unwittingly, to their offspring. These outcomes seemed all the more surprising in view of the considerable overall aggregate congruence between the two generations.Somewhat lost in the (over) generalizations flowing out of this and related research were a number of important qualifications. Transmission rates tended to vary in a systematic fashion according to type of political trait. The more concrete, affect-laden, and central the object in question, the more successful was the transmission. More abstract, ephemeral, and historically conditioned attributes were much less successfully passed on. Salience of the political object for the parents was an important conditioner of successful reproduction, as was perceptual accuracy on the part of the child (Acock and Bengston 1980;Percheron and Jennings 1981;Tedin 1980;Westholm 1999). The presence of politically homogeneous parents, and other agents allied with the parents, enhanced the fidelity of transmission (Jennings and Niemi 1974, chap. 6;Tedin 1980). Contextual properties such as larger opinion climates (Jennings and Niemi 1974, 81-82, 161-62) and party systems (Westholm and Niemi 1992) also affected within-family consonance. These specifications and qualifications also lent support to social learning theory explanations of how children come to resemble their parents more in some respects than others. Although not in the tradition of the transmission model, but fully compatible with social learning theory, other inquiries have revealed the importance of communication patterns within the family in shaping the political make-up of the child (e.g., Tims 1986; Valentino and Sears 1998).In this paper we return to the topic of intergener...