This study assessed whether and how self-explanation reading training, provided by iSTART (Interactive Strategy Training for Active Reading and Thinking), improves the effectiveness of comprehension processes. iSTART teaches students how to self-explain and which strategies will most effectively aid comprehension from moment-to-moment. We used RSAT (Reading Strategy Assessment Tool) to assess how iSTART changes the relation between important self-explanation reading strategies--bridging and elaboration--and online comprehension, and how often they are produced. College and high school students received iSTART and were administered RSAT prior to and post-training. Results from three experiments showed that iSTART primarily benefits bridging inferences when self-explaining. The frequency of bridging inferences was higher post training than prior to training, but only in the experiments involving college students. Additionally, prior to exposure to iSTART, RSAT bridging scores did not predict comprehension performance, whereas they did after iSTART, suggesting that iSTART may improve comprehension processes by teaching students how to appropriately use self-explanation to address comprehension difficulties. Finally, the results from this study suggest that RSAT may provide a valuable computer-based assessment of the effectiveness of self-explanations that could be used in conjunction with iSTART and in future research on self-explanation. (Contains 8 tables and 1 footnote.)