The effects of various questioning strategies on reading comprehension were tested. The subjects, 24 second grade students reading at or slightly above their grade level, were divided into three instructional groups. The strategy group focused on integrating textual material with prior knowledge. The question group got a steady diet of inferential questions. The control group received a traditional mix of literal/inferential probes (about a four to one ratio). In general, posttest ANOVA and pre/posttest ANCOVA supported all five of the study's hypotheses: (1) helping children make connections between what they already knew and what was in a text increased the likelihood that they would draw inferences spontaneously; (2) giving children practice in answering inference questions also tended to enhance their ability and inclination to do so spontaneously; (3) students trained in an inferencing set, whether by suggestion or sheer practice, processed explicit messages from the text better than students who focused on those messages; (4) treatment effects appeared to be subject to a prior-knowledge filter; and (5) growth in inferencing ability did not necessarily overcome the inherent advantage typically attributed to literal questions. Although the design of the study was not ideal and the data supported some hypotheses to a greater degree than others, it was concluded that instruction and practice in making inferences had direct consequences on children's reading comprehension. (RL)