BACKGROUND AND OBJECTIVE: Pediatricians are encouraged to engage in community child health activities, yet practice constraints and personal factors may limit involvement. The objective was to compare community involvement in 2004 and 2010 and factors associated with participation in the past year. METHODS: Analysis of 2 national mailed surveys of pediatricians (2004: n = 881; response rate of 58%; 2010: n = 820; response rate of 60%). Respondents reported personal characteristics (age, gender, marital status, child ≤5 years old, underrepresented in medicine), practice characteristics (type, setting, full-time status, time spent in general pediatrics), formal community pediatrics training, and community pediatrics involvement and related perspectives. We used χ2 statistics to measure associations of personal and practice characteristics, previous training, and perspectives with involvement in the past 12 months. Logistic regression assessed independent contributions. RESULTS: Fewer pediatricians were involved in community child health in 2010 (45.1% in 2004 vs. 39.9% in 2010) with a higher percentage participating as volunteers (79.5% vs. 85.8%; both P = .03). In 2010, fewer reported formal training at any time (56.1% vs. 42.9%), although more reported training specifically in residency (22.0% vs. 28.4%; both P < .05). Factors associated with participation in 2010 included older age, not having children ≤5 years old, practice in rural settings, practice type, training, and feeling moderately/very responsible for child health. In adjusted models, older age, practice setting and type, feeling responsible, and training were associated with involvement (P < .05). CONCLUSIONS: Formal training is associated with community child health involvement. Efforts are needed to understand how content, delivery, and timing of training influence involvement.