It has been hypothesized that students with low self-efficacy will struggle with complex reading tasks in assessment situations. In this study we examined whether perceived reading self-efficacy and reading task value uniquely predicted reading comprehension scores in two different item formats in a sample of fifth-grade students. Results showed that, after controlling for variance associated with word reading ability, listening comprehension, and nonverbal ability through hierarchical multiple regression analysis, reading self-efficacy was a significant positive predictor of reading comprehension scores. For students with low self-efficacy in reading, reading self-efficacy was a significant positive predictor of multiple-choice comprehension scores but not of constructed-response comprehension scores. For students with high self-efficacy in reading, reading self-efficacy did not account for additional variance in either item format. The implication that the multiple-choice format magnifies the impact of self-efficacy in assessments of reading comprehension is discussed. (Contains 4 tables, 2 figures, and 1 note.)