Social benefits and quality of group daycare vary greatly across countries; the effects on children may depend on the social context. This study used the prospective Norwegian cohort of 24,259 (MoBa; Norwegian Mother and Child Cohort Study) to examine predictors of group daycare, and the effects of family risk and group daycare on child behaviour at 36 months. The results showed that virtually all 9-month-olds remained in maternal care; while at 18 months over two-thirds attended some form of group daycare. High maternal education and family income very strongly predicted children's attendance in group daycare--a social selection opposite to that found in North America. Group daycare had a statistically significant effect on both distress and aggression, but the effect was very weak. Family risk was associated with both distress and aggression, but this did not differ according to type of care. The findings show a marked difference from those in North America, suggesting that social selection worked in the opposite direction. This was probably due to Norway's generous paid maternal leave and high-quality daycare. (author abstract)