This study represents a quasi-experimental test of the role of early social-emotional experience and adult-child relationships in the development of typically developing children and those with disabilities birth to 4 years of age living in orphanages in St. Petersburg, Russian Federation. The three orphanages in the current study were selected because they were among the best in St. Petersburg and their directors were willing to cooperate with the project. They met reasonable institutional standards for medical care, nutrition, physical environment, sanitation, toys and equipment, specialized professional services, and the lack of abuse, but similar to many other orphanages in the literature they were deficient in the social-emotional caregiver-child interactions and lack of relationship experiences provided to children. Caregivers performed routine duties in a perfunctory, business-like manner with minimum interaction with children. In addition, children had 9-12 caregivers per week, as many as 60-100 different caregivers over the first 2 years of life, and no caregiver today as yesterday or tomorrow. Two interventions designed to improve the social-emotional-relationship experience of children were implemented. "Training" used a train-the-trainer approach to educate staff on all aspects of early childhood development and mental health, emphasizing warm, caring, sensitive, responsive, and developmentally appropriate interactions especially during routine caregiving duties. "Structural changes" consisted mainly of reducing group size from approximately 12 to 6, assigning two primary caregivers to each subgroup so that a primary caregiver was available every day, terminating periodic transitions of children to new wards and caregivers, integrating groups by age and disability status, and establishing Family Hour for 1 hr in the morning and afternoon in which caregivers were instructed to be with their children. Thus, the interventions promoted a social-emotional behavioral style of interaction with children (e.g., responsiveness, talking, playing, shared emotions), not specific actions or programs of activities, plus an employment and organizational structure that provided more consistent contact with fewer caregivers and smaller age- and disability-integrated groups to facilitate caregiver-child relationships. The Baby Homes were nonrandomly assigned to one of three conditions: both training and structural changes (T+SC) were implemented in one orphanage, training only (TO) in a second, and no intervention (NoI) in a third. Results showed that the interventions were successfully implemented, and caregivers improved their ward behavior (HOME Inventory) over the course of the study, more in T+SC than TO. The interventions produced substantial improvements in the development of children, both typically developing and those with disabilities, greater for children in T+SC than TO than NoI, and greater after 9+ than 4-9 months of exposure to the intervention (i.e., dose-response and/or age effect). Developmental improvements occurred for physical growth (e.g., height, weight, chest circumference), all aspects of general behavioral development (Battelle Developmental Inventory), a variety of social-emotional-relationship behaviors (Parent-Child Early Relationship Assessment, Infant Affect Manual), and attachment categories, behavior ratings, and dimensions. Scientifically, these results are consistent with the frequently voiced hypothesis that the major corrosive element of early institutional experience is the relative lack of social-emotional experience and warm, caring, sensitive, responsive caregiver-child interactions and relationships, and they suggest the potential role such early experience plays in many aspects of development. From a practical standpoint, training was not very effective by itself; it was more beneficial in combination with structural changes that supported social-emotional relationships. These results provide a rationale for making similar improvements in other institutions and perhaps in foster care and nonresidential care environments as well and for balancing skill building with social-emotional-relationship training in early childhood personnel preparation curricula. (Contains 40 figures, 30 tables and 14 notes.)