This quasi-experimental research study examines whether the use of Assessment and Learning in Knowledge Spaces (ALEKS), an ITS, shows a statistically significant improvement in students' mathematics achievement than traditional teacher-led instructions. This non-randomized research study measured the efficacy of ALEKS on ''underachieving students' mathematics achievement among 158 (60 in teacher-led group and 98 in ALEKS-led group) 8th-grade students. A pretest and posttest were employed between teacher-led instructions versus ALEKS-led instructions from two consecutive years. During the first year, only McGraw's curriculum "Reveal" was used with no use of ALEKS. In the second year, the school implemented ALEKS as a supplemental tool in a mathematics support class for fifty minutes every other day for a year to provide instruction to struggling students along with McGraw's curriculum "Reveal." We also compare the results of five years of End of Grade (EOG) without ALEKS with one-year EOG with the use of ALEKS. Data were analyzed using paired t-test and analysis of covariance (ANOVA) to evaluate the efficacy of ALEKS on students' mathematics achievement. We find that the results of ALEKS-led and teacher-led instructions are highly statistically significant. The results show that teacher-led instructions are more effective because of higher test scores and lower variance for teacher-led instructions.