This study assessed what motor skill tests were appropriate by observing 636 young children's behavior in daily life (low-aged classes, M = 3.7 yr.; middle-aged classes, M = 4.7 yr.; high-aged classes, M = 5.7 yr.). A homeroom teacher and an assistant teacher estimated motor achievement as pass-or-fail, and then judged pass-or-fail based on practical testing. Estimate-re-estimate agreement, interrater agreement, and agreement between estimated values and measured values were examined for 27 items, e.g., skipping, bouncing a ball, and turning on one leg. Estimate-re-estimate agreement was high on the whole. Interrater agreement ranged from 34% to 100% for 3-yr.-olds, 21% to 100% for 4-yr.-olds, and 89% to 100% for 5-yr.-olds. Agreement between estimated values and measured values greater than 80% was found in most items for 5-yr.olds (14 items). After examining the above-mentioned agreements, 26 items were selected as possible tests to judge motor development, using a pass-or-fail, from the observation of young children's behavior in daily life.