This study aims to report the development an instrument that is standardized (reviewed by validity, reliability, and difficulty index) to detect science misconception in an elementary school teacher. This study used a 4-D model; defining, designing, developing, and disseminating. First, it was prepared with 47 opened-ended questions, and then it was validated by two experts (Physics and Biology) to find out content validity value and inter-rater reliability. The next stage was arranged an empirical test to 103 elementary school teachers in Surakarta, Indonesia by stratified purposive sampling. The data were analyzed using content validity formula, a measure of Agreement Kappa, Pearson's Correlation, Cronbach's alpha, and difficulty index. The result of this study indicated that there were 22 valid items. It can be concluded that this instrument is worthy to be used to detect the science misconception of an elementary school teacher.