This study aimed to measure the diagnostic accuracy of mathematical concepts and applications curriculum based measures (MCAP-CBM) in screening for fourth grade students at risk for mathematical learning disabilities. The study also sought to determine the benchmarks for the classification of students in MCAP-CBM, namely frustration, instructional., and mastery levels. To achieve the objectives of the study three alternative and adequate forms of MCAP-CBM were developed for screening and progress monitoring purposes. The measures included all the skills that were likely to be mastered by the end of the school year. The MCAP-CBM consisted of 24 items as described in the traditionally and empirically developed measures in the literature. The study sample included 393 students from both genders enrolled in fourth grade in schools in Muscat. The MCAP-CBM administered three times per the academic year in order to develop benchmarks for the classification levels. The 25th. And 75th. Percentile benchmarks were used to validate the benchmarks for the three classification level: Frustration, instructional, and mastery. T- test showed that there were statistically significant in frustration level in the favor of the female pupils. The results also showed no differences in the classification levels instructional and mastery) between males and females. The Receiver Operating Characteristic (ROC) Curve was used to examine the diagnostic accuracy of the MCAP-CBM using mathematics achievement scores as the outcome measure. The study presents a number of recommendations: curriculum based measurement can predict the performance of students and therefore it can be an effective element in decision making; it also helps in universal screening; and that the percentile 25th is a reliable benchmark for identifying students at risk for mathematical learning disabilities.