This study examined associations of preschool type (i.e., urban and suburban Head Start and university-affiliated center) and teacher-child variables with positive and negative child outcomes among 145 preschoolers (74 boys). Differences emerged across preschools, with urban Head Start children scoring lowest on the emotional competence measures and university-affiliated preschoolers experiencing less peer victimization than urban and suburban Head Start preschoolers. Differences across preschool types were also found for the teacher-child variables, such that teacher-child closeness was lower and teacher-child conflict and dependence were highest in the urban Head Start preschool. Regression analyses revealed significant and meaningful interactions between preschool type and teacher-child relational quality in the prediction of children's social-emotional outcomes. Teacher-child conflict was negatively associated with emotion regulation and teacher-child dependence was associated with the highest levels of emotion regulation, but only for university-affiliated preschoolers. Suburban Head Start preschoolers experienced less prosocial attention than urban Head Start preschoolers, but only when teacher-child closeness was high. Teacher-child closeness was also a negative predictor of urban Head Start preschoolers' prosocial attention. (author abstract)