To identify factors associated with children's motor skills. Cross-sectional. Australian preschool-aged children were recruited in 2009 as part of a larger study. Parent proxy-report of child factors (age, sex, parent perception of child skill, participation in unstructured and structured activity), self-report of parent factors (confidence in their own skills to support child's activity, parent-child physical activity interaction, parent physical activity) and perceived environmental factors (play space visits, equipment at home) were collected. Moderate to vigorous physical activity (MVPA) (ActiGraph GT1M accelerometer) and motor skills (Test of Gross Motor Development-2) were also assessed. After age adjustment, variables were checked for association with raw object control and locomotor scores. Variables with associations of p < 0.20 were entered into two multiple regression models with locomotor/object control as respective outcome variables. Motor skills were assessed for 76 children (42 female), mean [SD] age-4.1 [0.68]: 71 completed parent proxy-report and 53 had valid MVPA data. Child age, swimming lessons, and home equipment were positively associated explaining 20% of locomotor skill variance, but only age was significant (β = 0.36, p = 0.002). Child age and sex, unstructured activity participation, MVPA%, parent confidence, home equipment (all positively associated), and dance participation (inversely associated) explained 32% object control variance. But only age (β = 0.67, p < 0.0001), MVPA% (β = 0.37, p = 0.038) and no dance (β = -0.34, p = 0.028) were significant. Motor skill correlates differ according to skill category and are context specific with child level correlates appearing more important.