Two experiments were conducted to explore how repeated testing affects the pattern and the accuracy of the judgment of learning (JOL). Three learning conditions, repeated study, repeated test, and standard study-test alternation, and two retention intervals, immediate and one-day delay, were manipulated in both experiments. Pair-association learning was used in experiment 1, and text recall was tested in Experiment 2. In both experiments, JOL monotonically increased between sessions in the repeated study condition, whereas JOL decreased sharply in the first test session in the repeated test and the standard conditions. However, the correlation between the JOL and the performance in the final test was higher in the repeated test and the standard conditions. The difference in the pattern and the accuracy of JOL between the learning conditions suggested that testing not only facilitates learning, but also increase the accuracy of JOL because testing provides information about the retrieval fluency.