The effect of selected aspects of sport involvement on children's friendship expectations (FEs) was investigated by having 80 sport-involved children, grouped equally by gender from ages 9 through 12, complete a 40-item Likert questionnaire. The questionnaire contained eight FEs adapted to each of five discrete sport contexts reflecting team-sport involvement, team-sport non-involvement, same-team membership, opposing-team membership, and poorer players. MANOVAs showed that the sport context was the principal effect. Posttesting revealed that the children agreed that team sport and same-team membership promotes friendship relations. They were relatively undecided whether non-involvement, opposite-team membership, or lack of skill interferes with friendship relations, although they agreed that the poorer player has more friendship problems in sport. Age, sex, and FE item interactions were comparatively small. Older children were more tolerant of the effects of opposing-team membership, older girls were more tolerant of lack of skill, and FE contrasts between sport contexts had good construct validity.