OBJECTIVE: The aim of this study was to use family-centered measures to estimate the effect of a collaborative quality improvement program designed to help practices implement systems to promote early childhood development services. METHODS: A cohort study was conducted in pediatric and family practices in Vermont and North Carolina. Eighteen collaborative education practices and 17 comparison practices participated in a 12-month program to assist practices in implementing improved systems to provide anticipatory guidance and parental education. The main outcome measures were change over time in parent-reported measures of whether children received each of 4 aspects of recommended care, documentation of developmental and psychosocial screening, and practice-reported care delivery systems. RESULTS: The number of care delivery systems increased from a mean of 12.9 to 19.4 of 27 in collaborative practices and remained the same in comparison practices (P=.0002). The proportion of children with documented developmental and psychosocial screening among intervention practices increased from 78% to 88% (P<.001) and from 22% to 29% (P=.002), respectively. Compared with control practices, there was a trend toward improvement in the proportion of parents who reported receiving at least 3 of 4 areas of care. CONCLUSION: The learning collaborative was associated with an increase in the number of practice-based systems and tools designed to elicit and address parents' concerns about their child's behavior and development and a modest improvement in parent-reported measures of the quality of care.