This study aims to compare Psychometric properties to examine the numerical ability for Gulf Cooperation Council Countries, according to the item response theory. To ensure achieving the objective of the study, the researcher used the data of numerical ability test in Gulf measurement for various mental abilities of the first and second primary grades, specifically in the sex Gulf Cooperation Council Countries; (UAE, Bahrain, Saudi Arabia, Oman, Qatar and Kuwait). This measurement contains thirty items, and the sample size contained of (4206), and the sample sizes of the GCC countries also converged. The three dimensional logistics model (3PLM) has been utilized with using MULTILOG 7.03 program to estimate the items features and overall abilities of students in the total sample and the calibrated samples. The results showed the stability of test items features estimations in general as; (discrimination, difficulty, valuation), and it must be mentioned that difficulty features had the highest level of stability. Additionally, the results showed that the frequency distribution of the students' abilities in the test was stable by the difference of calibration samples. Furthermore, the average amount of information for the calibrated samples was similar to the total sample, while the amount of information and the location of the maximum values through the connected ability were different in each sample of the calibration in the GCC Countries, and the information amount enhanced as the participant ability increase to reach the highest range at the average abilities. Whereas this range deceases at the two ends of the capacity diagram at each sample of GCC calibration samples. The researcher recommended the possibility of relying on the test of numerical capacity of the GCC countries internally in different applications because of the demonstrated degree of stability and accuracy in assessing the capabilities of individuals.