The purpose of this study is to identify the effect of demonstration type of model on the Volleyball ability of elementary school students. Subjects were divided 48 6th grade girls who have never played the game before into four groups randomly assigned 12 subjects : A mate model observation group, A mate skilled model observation group, A teacher model observation group, A teacher silled model observation group and each group observed demonstration of model and therefore analyzed the result of performance on acquisition phase and retention phase in blocked practice groups and practice groups. Every two blocked practice group performed 5 trials without demonstration of model in the retention phase 24 hours after each blocked practive group performed 10 trials ( practive two blocked a day ) for 3 days in the acquisition phase. The ecperiments was conducted by two-way ANOVA and factorized as independent variable with 4 practice groups and 6 bolcked practice groups for 3 day 4 x 6 in the acquisition phase and with 4 practice groups and 2 blocked practice groups 4 x 2 in the retention phase. Statistical analysis put to use SPSS/PC 10.0 program. The results of this study are as follows: 1) In the Acquisition and Retention phase of the Volleyball ability learning, a mate skilled model observation group were more effective than a mate model observation group. 2) In the Acquisition and Retention phase of the Volleyball ability learning, a teacher skilled model observation group were more effective than a teacher model observation group. 3) In the REtention Phase of the Volleyball ability learning, a mate model observation group were more effective than a techer model observation group. Therefore, it make the best use of demostration of model not only a mate skilled but also a teacher skilled in elementary physical calss.