This study examined the basic measurement theories (classical test and item response theories) and psychometric properties of the adopted 2016 - 2017 National Examination Council (NECO) mathematics examination instruments in Calabar Education Zone of Cross River State, Nigeria. The research design adopted for the study was the instrumentation design. The stratified random sampling technique was used to select a sample of one thousand three hundred and fifty one (1,351) students across the local government area in the zone. Two instruments, namely the 2016 and 2017 NECO mathematics multiple choice tests were used for data collection. A trial test of the instruments on 60 comparable non sample group showed reliability (KR20) indices of .91 and .94 for the 2016 and 2017 versions respectively. Difficulty indices ranged between .20 and .85 as well as .10 and .91 respectively for the two years. Discrimination indices were .42<d< 1.00 and .31< d< .89 respectively for the two versions. The final administration yielded data that were collated and analyzed using 2-way repeated measure ANOVA and chi-square statistics. Statistical analysis, showed that the item difficulty and item location, item discrimination indices and ‘a’ parameter and the option distraction powers of the 2016 and 2017 NECO mathematics multiple choice tests were significantly different. It was concluded that significant difference exist between psychometric properties using the different measurement theories. Based on these findings, it was recommended that NECO should use the item response theory (IRT) in assessment for its relative merit over the CTT among others.