Purpose Preschoolers born preterm are at an increased risk for the development of language impairments. The primary objective of this study was to document the expressive language skills of preschoolers born preterm through 2 assessment procedures, language sample analysis, and standardized assessment. A secondary objective was to investigate the role of nonlinguistic factors in standardized assessment scores. Method The language skills of 29 children born preterm (mean gestational age of 31 weeks) were compared to a group of 29 preschoolers born full term. Language samples were collected during free play and objective measures of semantic and grammatical skills were calculated. Likewise, grammatical and semantic measures of language were obtained from a standardized assessment. Information on nonlinguistic factors, including attention, hyperactivity, and nonverbal intelligence quotient, was also collected. Results The results of analyses of variance indicated that the children in the PT group had significantly poorer performance than the children born full term on all of the measures of language skill obtained from the language sample analysis. Group differences were found for only 1 measure of language skill obtained from the standardized assessments. Nonverbal factors were not found to account for group differences in assessment scores. Conclusions Generally, the children born preterm performed more poorly when language skill was measured via language sample analysis than standardized assessment. These findings underscore the importance of using language sample analysis in conjunction with standardized assessment in the evaluation of children's developing language skills.