The authors examined whether relevance instructions compensate for differences in verbal ability on measures of reading time, text recall, and sentence recognition. College students (n = 81) with higher and lower verbal ability were assigned randomly to 1 of 2 relevance-instruction conditions before reading a text. They asked participants in each condition to focus on different categories of information within the same text. Relevant information took longer to read and was recalled and recognized better than nonrelevant information. Readers with higher verbal ability read faster and recalled and recognized more information correctly than did those readers with lower verbal ability. Results support the noncompensatory hypothesis, which states that relevance instructions and verbal ability make independent contributions to resource allocation and learning. Readers with lower verbal ability may need additional support even when given prereading relevance instructions. (Contains 3 tables and 1 figure.)