The purpose of this study was to investigate the effect of instruction and multiple intelligence in order to develop student’s superior ability and to avoid meeting student’s weakness instead of improving IEP instruction planning and traditional remedial inefficience. The main sample distributed to experimental group and control group. The participants includes second graders with handicapped, parents or guardian of the child with handicapped, regular classroom teacher, resource room teacher, reseeding in resource room of the elementary school in National Ping-Tung Teachers College and Tsau-Zou elementary school for one year experimental study. In addition, the investigator as associate professor in National Ping-Tung Teachers College and Tsau-Zou elementary school for one year experimental study. In addition, the investigator as associate professor in National Ping-Tung Teachers College observes the whole experimental process and the classified the effect of each participant on the basis of Multiple Intelligence Checklist, self-evaluation achievement score, the checklist of learning behavior characteristics dual conversation by recorder, and the observation by camera. Data analysis used as two-way design ANOVA to compare with pre-test and post test difference in the experimental group and control group. Discussed the experimental results and the observation for special education teachers and future researchers. The main results of this study as follows: Between experimental and control group in self-evaluation multiple intelligence has no significant difference. Experimental group at the logic-mathematics intelligence, vision-space intelligence, the body-movement intelligence, music rhythm intelligence and the whole intelligence in post-test score is higher than pre-test score. But, control group has not any significant difference. The experimental group in literature and mathematics intelligence both has higher score than the control group on the basis of individualized education program. The control group has higher score in mathematics achievement test in post-test than the experimental group on the basis of academic performance. Teachers’ parents, administrators’ attitudes toward multiple intelligence in both groups has no significant difference in pre-test and post-test.