Because reading achievement tests are often used for student assessment and placement, it is important to understand how children of limited English proficiency perform on such tests, and how their performance is related to their literacy development. In this study, both quantitative and qualitative methodologies were employed in order to identify factors that influenced the English reading test performance of 51 Hispanic children, as compared with the performance of 53 Anglo children enrolled in the same fifth- and sixth-grade classrooms. The children's reading test performance was examined for differential effects of time constraints, the children's prior knowledge, and question type (explicitness of the information requested). Differences in the children's general and test-specific vocabulary knowledge were also noted. In addition, 18 children participated in retrospective, open-ended interviews that focused on how they determined their answers to the vocabulary and reading tests. The combined findings suggest that the Hispanic students' reading test scores seriously underestimate their reading comprehension potential. Their test performance was adversely affected by their limited prior knowledge of certain test topics, their poor performance on the scriptally implicit questions (which required use of background knowledge), their unfamiliarity with vocabulary terms used in the test questions and answer choices, and their tendency to interpret the test literally when determining their answers. When differences in prior knowledge were controlled statistically, the overall reading performance of the two groups did not differ. Giving the children more time to complete the test did not help the Hispanic children's relative performance because both groups' performance improved similarly. The Hispanic children's interview responses tended to elicit more information about their test passage comprehension than did their test performance.