To provide researchers with an instrument, valid and reliable enough, for measuring students’ metacognitive knowledge needed in solving mathematical word problems, based on the theoretical foundation and previous research, a set of 24 questions at three levels of metacognitive knowledge was constructed. The initial validity of these questions was confirmed by Psychology Professors and high school departmental heads in math teaching. To measure the basic qualities of validity and reliability of the constructed instrument, it was used in interviewing 486 high school students who were selected from among all such students in two school districts in Tehran using cluster sampling method. The recorded verbal responses of the participants were then coded and analyzed. Results confirmed the criterion validity of the instrument. As for its reliability, a test-retest of the same group yielded a correlation of 0.69 and an inter-rater reliability was 0.99. The internal consistency of the questionnaire was also assessed to be 0.73. This questionnaire is recommended to be used with the 13-15 years old age group.