The purpose of this study was to examine the relationship between a teacher report measure, the Teacher Rating of Oral Language and Literacy (TROLL; Dickinson et al. in "Teacher rating of oral language and literacy (TROLL): a research-based tool." Center for the Improvement of Early Reading Achievement, University of Michigan, Ann Arbor, 2001) and a direct behavioral measure of language development, the Preschool Language Scale-4 (PLS-4; Zimmerman et al. in "Preschool Language Scale-4." The Psychological Corporation, San Antonio, 2002), among English-speaking (n = 210), Spanish-speaking (n = 34), and English/Spanish bilingual (n = 109) typically-developing preschool children. Three hundred and fifty-three preschool children who attended early childhood education programs in an urban area of the Southwestern United States participated. Preschool teachers completed the TROLL, and the PLS-4 was individually administered to the children at preschool centers. The TROLL and PLS-4 were significantly correlated for English-speaking children, but with small effect sizes noted. For Spanish-speaking children, the TROLL and the expressive subscale of the PLS-4 were not significant, and for bilingual children the TROLL and PLS-4 were not significant. English-speaking children scored higher on the TROLL than the Spanish-speaking and bilingual children. Finally, a higher proportion of Spanish-speaking and bilingual children received a TROLL score at or below the 10th percentile. Results suggest that the TROLL did not adequately capture typically developing children's linguistic and literacy development in a uniform manner across language groups. Caution is recommended when relying upon a single instrument to describe the emergent literacy and language skills of preschool children from Spanish-speaking and bilingual backgrounds.