This longitudinal study aimed to examine how physical activity parenting (PAP) directly predicted objectively measured children’s moderate-to-vigorous physical activity (MVPA) and sedentariness over a three-year transitional period from early to middle childhood, and second, whether the children’s perception of motor competence (PMC) mediated or moderated the influence of PAP to children’s MVPA or sedentariness. At time 1 (T1), PAP and children’s (N = 396, mean age 5.80, SD 1.04) PA were assessed by parental questionnaire. Three years later, at time 2 (T2), children’s (N = 396, mean age 8.80, SD 1.04) PMC was measured by a validated pictorial scale, and MVPA and sedentariness were measured by accelerometers. All the analyses were conducted using the Mplus statistical package (Version 8.4). The models were adjusted for the following covariates: children’s PA (T1), gender (T1), age (T1), mean accelerometer measurement in hours per day (T2), and parents’ education level (T1). Results showed that PAP at T1 did not significantly predict level of MVPA or sedentary time at T2 and, therefore, PMC did not mediate the PAP-children’s MVPA or sedentary time relationship either. However, PMC significantly moderated the relationship between PAP and MVPA but not between PAP and sedentary time. The results suggested that parental support positively predicts children’s MVPA among children with low PMC but not among children with high PMC. This unique finding proposes that family-based PA interventions could benefit from screening of children with low PMC and provision of PA counselling to their parents.