People read for many different reasons. These goals affect the cognitive processes and strategies they use during reading. Understanding "how" reading goals exert their effects requires investigation of whether and how they affect specific component processes, such as validation. We investigated the effects of reading goal on text-based and knowledge-based validation processes during reading and on the resulting offline mental representation. We employed a self-paced sentence-by-sentence contradiction paradigm with versions of texts containing target sentences that varied systematically in congruency with prior text and accuracy with background knowledge. Participants were instructed to read for general comprehension or for study. Memory for text information was assessed the next day. We also measured the degree to which each text topic was novel to a reader, as well as his or her working memory capacity. Results show that reading goals affect readers' general processing as indicated by overall reading times, but provide no evidence that they influence validation processes. Reading goals did affect readers' memory for target information but this effect depended on congruency between that information and the preceding text: Reading for study generally resulted in better memory for target information than reading for comprehension did, but not for target information that was incongruent with prior text. These results suggest that reading goals may not influence validation processes directly but affect subsequent representation-building processes after the detection of an (in)consistency--particularly in the case of incongruencies with prior text.