Introduction: Vocabulary performance may vary due to environmental stimuli, and socioeconomic and cultural context. Purpose: To compare the performance of expressive vocabulary of children attending public and private schools, aged from four to five years, and analyze the most frequent substitution processes. Methods: The sample consisted of 86 preschool children, aged from four to five years, male and female, without signs of syndromes and neurological or language disorders, divided into G1: Group of preschoolers in the Public School, and G2: Group of preschoolers in the Private School. All children were submitted to speech-language screening and expressive vocabulary evaluation through the ABFW Vocabulary Test. Data were submitted to statistical analysis, using Fisher's Exact Test and T-test (p <0.05), considering the occurrences of Designations by Usual Word, Non-Designations and Substitution Processes according to conceptual fields. Results: G2 presented better performance in the vocabulary test than G1, with statistically significant difference for all conceptual fields. However, when comparing pre-school performance with test parameters, most G1 and G2 preschoolers presented adequate performance. The most frequent substitution processes were those of hyperonymy and close co-hyponymy. Conclusion: Private school preschoolers present better performance in expressive vocabulary tests. In addition, substitution processes occurred more frequently in preschoolers in the public school system, and the most frequent ones were hyperonymy and co-hyponymy.