Category learning is defined as the process of acquiring knowledge about categories so that it can be subsequently used during categorization. The aim of the study was to investigate metacognitive monitoring in different learning stages during the acquisition of category learning tasks with two mutually exclusive categories. Rule-based category learning tasks which are expected to activate the explicit learning system, and integration-information category learning tasks which are expected to activate the implicit learning system were examined. It is assumed that the system activated during category learning affects the ability of metacognitive monitoring of the ongoing learning process and subsequent categorization. Metacognitive monitoring has been examined by confidence judgments, feeling-of-warmth judgments, judgments of learning and category learning judgments. The results showed that classification accuracy and metacognitive judgments are affected by task type. When the explicit learning system is activated, classification accuracy and metacognitive judgments are high and compatible. However, both measures decrease as a function of increasing complexity of the logical rules underlying the tasks. When the implicit learning system is activated, classification accuracy and metacognitive judgments are lower and less compatible. Furthermore, the results showed that in the final learning stage the sudden increase in classification accuracy directly before complete accuracy is achieved, is followed by gradual increase in metacognitive judgments which suggests parallel activation of both, explicit and implicit, category learning systems. To conclude, the results showed that metacognitive judgments in the category learning domain are formed based on different available cues depending on which category learning system is activated during learning.