The purpose of this study was described the interrater and then-rest reliability of the gross motor of the Peabody Development Motor Scale-2(PDMS-2). Subject were 33 children without delay development  between 26 months and 35 months. Three physical  therapists viewed videotapes of the children and scored each child's performance on the gross motor scale of the test. 33 children were tested on one occasion, with each children scored independently and simultaneously by three physical therapists to examination interrater reliability. 30 children were each by two therapists to determine test-retest reliability within two weeks. The ANOVA statistics allows examination of the magnitude of the different sources of variability. The interrater reliabilities were analyzed for al three testers with ICC. To calculated significant differences between test and retest, paired t-test was used. The test-retest reliabilities were calculated with the ICC. The intraclass correlation coefficient(ICC) for interrater and test-retest reliability of measurements obtained with the PDMS-2 (GM) were high(ICC=0.89 ~ 0.97). The ICCs for inter-rater reliability varied from 0.92 to 0.97 for three sub-test scores and 0.93 for gross motor quotient. The ICCs for test-retest reliability varied from 0.89 to 0.94 for three sub-test scores and 0.92 for gross motor quotient. The finding suggest the PDMS-2 to be a useful and reliable instrument for assessing gross motor function and identifying children with delayed or aberrant motor development in Korea.