Four experiments, each building on the results of the previous ones, explored the effects of several manipulations on learning and the accuracy of metacognitive judgments among deaf and hard-of-hearing (DHH) students. Experiment 1 examined learning and metacognitive accuracy from classroom lectures with or without prior "scaffolding" in the form of a description of main points and concepts. Experiment 2 compared the benefits of scaffolding when material was read versus when it was presented as a lecture signed for DHH students and spoken for hearing students. Experiment 3 compared scaffolding provided in the form of main points versus vocabulary, and Experiment 4 examined effects of material familiarity and a delay between study and test. Results indicated that although all students had a tendency to overestimate their performance, hearing students learned more and were more accurate in their metacognitive judgments than DHH students. Content familiarity improved the accuracy of metacognitive judgments by both DHH and hearing students, but the delay manipulation was effective only for hearing students. Consistent with other recent findings, DHH students learned as much from reading as they did from signed instruction. Differences between DHH and hearing students may indicate the need for explicit instruction for DHH students in academically relevant skills acquired incidentally by hearing students.