Rhyming tests have historically been used in the education system to assess reading readiness. English language learners (ELLs) have consistently scored poorly on these assessment tools. The current article examines a possible reason for this poor performance by ELLs. Specifically, the authors examined the relationship between semantic associations of visual images and performance on rhyming assessments for ELLs. Two groups of students, native English speakers and native Spanish speakers, were tested using a typical rhyming assessment tool. As expected, it was found that the native English speakers outperformed the native Spanish speakers. An analysis of the native Spanish speakers' errors revealed semantic interference. Educational implications of these findings are discussed. (Contains 3 figures and 1 endnote.)