This paper reports the findings of a quasi-experiment that aims to investigate the role of metacognitive teaching-learning (MTL) and mathematical prior knowledge (MPK) on students' mathematical logical thinking ability (MLTA) and self-regulated learning (SRL). The subject of the study was 70 tenth-grade students inSumedang, Indonesia. One experimental group and one control group were used, and the students were grouped into three categories based on their mathematical prior knowledge (high, medium and low). The instruments of this study were a short-answer-MPK test, an essay MLTA test, and a SRL scale. The data were analyzed by using Anova, t-test, Mann-Whitney test, and Kruskal-Wallis test. The study find that compared to conventional learning, MTL gave better influence on students with medium MPK in MLTA and gave an overall better effect on students in SRL. However, for entirely students, the students' grades of SRL are fairly good, but the students' grades of MLTA are still low. There is no interaction between learning approaches and MPK on improving MLTA, and there is an interaction between learning approaches and MPK on improving SRL.