This study aims at developing an instrument to measure metacognition in solving physics problems among college students. This study used Research and Development (R & D) model with the non-test instrument development model. The instrument was in the form of a questionnaire that consisted of four choices with the scale was ranging from 1 to 4. There were 664 first-year college students participating in this study which were chosen with purposive sampling technique. They had taken a basic physics course at Tadulako University. Before completing the questionnaire, the students were asked to solve the two physics problems. The expert judgment conclusions were analyzed by means of Aiken formula. The empirical data analysis employed the Quest program to test the Goodness of Fit (GoF) to Partial credit model (PCM) and item validity, while the instrument reliability and the analysis factors for the validity of the metacognition dimension in solving physics problems were tested with the assistance of the SPSS program. The results of the content validity with the Aiken's V value was .81 and the reliability was .90, and the mean of INFIT MNSQ 1.00 with a standard deviation of .18 respectively. It was identified that eight factors of metacognition that contributed to solve physics problems, such as declarative knowledge, procedural knowledge, conditional knowledge, planning, monitoring, evaluation, debugging and information management. These results can be used as the guidelines to measure students' metacognitive abilities in solving physics problems.