Abstract: Researchers in language learning divide a given learner’s passive vocabulary, or words that can be defined on request, from his or her active vocabulary, or words that are utilized in the production of natural language. Active vocabulary is of greater value for communicative competence and of more interest to language testers, but is harder to adequately assess than passive vocabulary. In the present study, the timed essays of writers from China, Korea, and Japan were assessed for their active vocabulary, operationalized as lexical diversity, using three popular metrics for such assessment. These measures were correlated with each writer’s score on a test of passive vocabulary, the VST, or vocabulary size test. Regression analysis was conducted for all three metrics as well. Across all lexical diversity metrics and language backgrounds, the correlation with VST score was low, suggesting that there is little direct connection between a writer’s passive vocabulary and their active vocabulary as expressed in their writing. This suggests that such tests are of little use for predicting practical vocabulary use in writing and should not be utilized for that purpose.