In this study, it was aimed to develop a valid and reliable number sense scale for elementary school students. Construct validity studies were conducted with 299 elementary school 4th grade students studying in 2 public schools in Eskişehir province selected by cluster sampling method. Criterion validity studies were conducted with 312 elementary school 4th grade students studying in 4 public schools in Eskişehir. As a result of the literature review and expert opinions, it was decided that the theoretical structure of the scale would consist of 6 factors and 31 items, namely the meaning and size of numbers, decomposing and combining numbers, determining reference points, the effect of operations on numbers, the flexible use of numbers, and the sensibility of numbers. Item analysis studies were first conducted on the draft scale and it was decided to remove 6 items with low discrimination from the test. Following the item analysis studies, construct validity studies were conducted. Within the scope of construct validity, it was aimed to test the theoretical structure created by the researchers. In this context, confirmatory factor analysis (CFA) technique was thought to be more appropriate for the purpose. As a result of CFA, it was determined that the theoretical structure of the scale produced a high fit with its 6factor 25-item structure. It was determined that the reliability coefficients of the whole scale and each factor were sufficient. Finally, within the scope of criterion validity, the relationship of the number sense scale with the problem solving achievement test developed by Ulu (2017) was examined. As a result, it was seen that all factors of the number sense scale had a moderately significant relationship with problem solving success, and this finding allowed us to see that the criterion validity of the number sense scale was sufficient.