Contemporary theories of learning suggest that successful learners are active in the learning process and that they tend to use a number of metacognitive processes to monitor learning and remembering. Drawing on the theoretical framework of Nelson and Narens (1992), the current study examined the effect of certain variables on metamemory processes and on students' ability to recall and recognise learned material. The present study explored the effect of four independent variables on five dependent variables. The independent variables were: 1. degree of learning (responses given until 2 or 8 times correct), 2. judgment of learning (JOL) timing (given immediately after learning session or 24 hours later), 3. retention interval between study and test (2 or 6 weeks), and 4. type of material studied (sentences, in or out of context). The dependent variables were: 1. judgement of learning (JOL), 2. confidence rating, 3. feeling of knowing (FOK), 4. recall, and 5. recognition.. As ancillary analyses, the study explored, firstly, whether gender differences had an effect on meta-level and object-level memory, and secondly, whether students who recalled more also made more accurate metamemory judgements. The effects of the independent variables on recall and recognition were consistent with those found by previous studies. The most interesting new finding of the present study was that students who made JOLs after twenty four hours were more likely to take into account the effect of the interval between learning and testing. Students who made immediate JOLs did not allow for the effect of the time interval on retention. A further new finding was that gender appeared to have had an influence on JOLs. The findings about the effects of timing of JOLs and of gender effects on JOL have implications for metacognitive theory and will stimulate further research. The practical significance of this research, particularly the implications for study skills training for all students, was that educators cannot presume that students will correctly predict what they will recall after six weeks if they make that judgement immediately after learning has occurred. Therefore, the effects of the passage of time on memory, and the efficacy of delaying judgments, should be made explicit. The finding that the manipulation of JOL timing has a significant effect on the accuracy of judgements has implications in the wider area of educational policymaking and for the current debate on competencies and quality assurance. Learning cannot be considered a simple process and when a large component of learning is selfdirected, as it is in tertiary institutions and increasingly in schools, many variables are operating.