The purpose of this study was to examine the effects of subitizing based number sense intervention on number sense and counting for students at-risk of MLD. Sixty students from two schools and a child welfare center took two kinds of screening tests: number sense and counting. The study involved three students with less than 15 percentile of the results who were not in exclusion criteria. This study adapted multiple-probe design across participants. I provided subitizing based number sense intervention to participants 40 min per session, 2-3 times per week, for eight weeks total of 21 sessions in their child welfare center or their classroom. As dependent measures, the formative test of BASA-EN and TEMI-AC used for number sense and counting also delayed post-test was conducted two weeks after intervention. The intervention fidelity was 97.08 percent, and the social validity was 95.2 percent for teachers and 91 percent for students. Intra-rater reliability was 99.91 percent to 100 percent. Data are graphed for each student with visual analysis including the effect factor: NAP, Tau-U. The pre and post percentile of standardized assessments also be calculated. The results as follows. The effect factor showed a significant positive effect in subitizing based number sense intervention on number sense and counting. Positive effects maintained delayed post-test. Post-test results of number sense standardized assessments improved from the fifth stage, which requires overall continuous intervention, to the third stage, which is the average level of performance. Post-test results of counting improved from below 15 percentile to the 32~45 percentile. These results indicate that subitizing based number sense intervention had a positive effect on the number sense and counting of elementary students those at risk of MLD. This paper discussed some implications, limitations, and suggestions for future research.