The influence factors of judgment of learning (JOL) have been the focus in this field for many years. Koriat (1997) proposed cue-utilization model to explain several cues that affect JOL, and pointed out the possible roles of these cues. Based on this model, Koriat and Bjork (2006) proposed a dual-processing theory of metacognitive monitoring, which suggested that metacognitive monitoring was an experience-based and theory-based dual-processing process. The former is an unconscious, non-analytic process, and the latter is a conscious, analytical process. The interaction model between these two processes, though, has been proposed since last century, it is difficult to directly manipulate the non-analysis process till now. In?the?early?stages?of JOL reseach, there were many studies focusing on theory-based analytical processing. Recently, more and more research has been concerned about the impact of processing fluency, emphasising the metacognitive experience on JOL. Some studies have verified the processing fluency effects of JOL in visual and auditory modalities. The motoric fluency effect has also been confirmed recently. Compared with visual and auditory fluency, motoric fluency is less influenced by beliefs, and more involved in non-analytic processing. This provides a new perspective for us to examine the non-analytic process. The purpose of the present study is to examine the effects of experience-based motoric fluency and theoretical-based word frequency on JOL and their possible interactions, exploring the roles of non-analytic processing and analytical processing in metacognitive monitoring. Some researchers believe that there may be interactions between conscious and non-conscious processes. To explore the role of non-analytic processing in memory monitoring, we further added cognitive load to weaken the possible advantage of analytical processing. ? In the present study, participants were asked to write words with their handedness or non-handedness to achieve different motoric fluency. And word frequency (high frequency/low frequency) was also manipulated in both experiments. Participants were required to learn words and to complete the JOL without (Experiment 1) or with cognitive load (Experiment 2). After a arithmetic distractor task, participants completed a 5-minute recall test. Results showed that: (1) when participants wrote words with their handedness, their JOLs were higher than those with their non-handedness, and JOLs on high-frequency words were higher than those on low-frequency ones. (2) The influence of motoric fluency on JOL changed due to the increasing of word frequency. Under the cognitive load condition, the influence of word frequency on JOL was different due to the change of motoric fluency. The results support the fluency effect and word frequency effect on JOL. More importantly, it suggests that analytical processing and non-analytic processing could work together in the process of memory monitoring. From the perspective of the dual-processing theory of metacognitive monitoring, the present study reveals the pattern of analytical and non-analytic processing on JOL, which plays an important role in exploring the mechanisms underlying JOL.