The purpose of this study conducted in 36 infant-toddler centres ("Kinderkrippen") in the city of Munich in Bavaria/Germany was to explore structural characteristics of early child care and education and their effects on child care quality. Stepwise regressions and variance analysis (Manova) examined the relation between quality of care and structural characteristics such as group size, staff competence and preparation time. A multi-respondent approach was employed, including self-report questionnaires of educators ("Erzieherinnen") as well as data on local working conditions collected from the directors of the centres. The findings of this study suggest that different sets of variables can be used to predict quality of education for under-threes. The practitioners' reported quality of education was found to relate to higher competence (with regard to centre programme, documentation, teamwork and staff qualification), better orientation towards the children's needs and better working atmosphere. Furthermore there were multivariate effects of the availability of and networking with external specialists, preparation time, and grouping arrangements within the centre such as changing from a home-group approach to an open-group setting. The findings are discussed in terms of their implications for research and the basic impact of education in early childhood. (Contains 4 tables and 1 figure.)