Background:Parental modeling has been shown to be important for school-aged children’s physical activity (PA) and television (TV) viewing, yet little is known about its impact for younger children. This study examined cross-sectional and 3-year longitudinal associations between PA and TV viewing behaviors of parents and their preschool children.Methods:In 2008–2009 (T1), parents in the Healthy Active Preschool and Primary Years (HAPPY) cohort study (n = 450) in Melbourne, Australia, self-reported their weekly PA and TV viewing and proxy-reported their partner’s PA and TV viewing and their 3- to 5-year-old preschool child’s TV viewing. Children’s PA was assessed via accelerometers. Repeat data collection occurred in 2011–2012 (T2).Results:Mothers’ and fathers’ PAs were associated with PA among preschool girls at T1, but not boys. Parents’ TV viewing times were significant correlates of girls’ and boys’ TV viewing at T1. Longitudinally, mothers’ PA at baseline predicted boys’ PA at T2, whereas sex-specific associations were found for TV viewing, with mothers’ and fathers’ TV viewing at T1 associated with girls’ and boys’ TV viewing respectively at T2.Conclusions:The PA and TV viewing of both parents are significantly associated with these behaviors in preschool children. The influence of the sex-matched parent appears to be important longitudinally for children’s TV viewing.