This study aimed to develop a standardised instrument for diagnosing science misconceptions in primary school children. Following a developmental research approach using the 4-D model (Define, Design, Develop, Disseminate), 100 four-tier multiple choice items were constructed. Content validity was established through expert evaluation by six science specialists. The study was conducted using stratified purposive sampling on 140 pupils from public and private primary schools with 'very good' and 'good' ratings in Surakarta, Indonesia. The instrument's construct validity, reliability, and empirical validity were assessed using Aiken's V, Pearson's Correlation Coefficient, and Cronbach's alpha, respectively. Difficulty index was also calculated. The final version comprised 61 valid and reliable items with Aiken's V ranging from 0.79 to 0.92, Pearson's Correlation ranging from 0.17 to 0.58, and a Cronbach's alpha coefficient of 0.86. Item difficulty ranged from 0.014 to 0.62, with 27 and 34 items falling into the moderate and difficult categories, respectively. This instrument demonstrates potential for effectively detecting science misconceptions in primary school learners. Based on these findings, teachers, policymakers and parents can take targeted action to address misconceptions. This may involve evaluating science textbooks, enhancing teacher training, and fostering discussions with science experts. Additionally, parental involvement in reinforcing science concepts is crucial.