Occupational therapists regard play as the primary occupation of childhood. However, many authors of play assessments have viewed play either as a functional outcome of other skills or as an indication of the child's developmental level. While such views of play are valid, these approaches to play are broad and do not fully consider the cognitive value of play. A study was undertaken with 82 preschoolers to see if a new play assessment which considers cognitive play skills, called the Child-initiated Pretend Play Assessment, (ChiPPA), possessed acceptable inter-rater reliability, and could discriminate between the play of typically developing preschoolers and preschoolers with pre-academic problems. This assessment evaluates preschoolers' elaborate imaginative actions, number of imitative actions, and number of object substitutions. The ChiPPA was found to have excellent inter-rater reliability and did discriminate between children who were typically developing and children who experienced pre-academic problems. In particular, the elaborateness of a child's play actions and the child's ability to substitute objects were strong discriminators of preschoolers with and without pre-academic problems. Three play profiles which emerged from the research are discussed.