This study consists of two parts. Part I The purpose is to make clear the consistent tendencies of individual differences of dependency at early childhood and also to study relation-ships of dependency and independency. Subjects: Thirty children from 15 months to 35 months old of highly educated parents. Method: (1) Observation in an experimental situation where mother and child may play freely for thirty minutes. Two investigators describe their behavior as it appears, which are analyzed later. (2) Questionaire on dependency, consisted of twelve items.(3) Time table of two successive days recorded and submitted by mother herself, based onbehaviors of both mother and child. Result: 1) There are the individual differences of dependency on each measure as shown on Table 1. 1, 1. 2, and 1. 3. 2) Correlations between dependency scores or indepedency scores by different methods are not so high as shown on Table 2. 1 and 2.2 (r=01-. 44). It was suggested that interviewing could be the more adequate method to know the overall situation. 3) Correlations between dependency and independency are not so hig h,(-. 10, -. 24, -. 59) as shown in Table 3. Part II The purpose is to study the relationshipof dependency, compliance and parent's attitude to child. Hypotheses: 1) Dependency drive will be reinforced if child's dependent behavior may be accepted by parents. 2) Strength of dependency drive will be related to the amount of punishment given by parents once after dependency drive is acquired. Dependency drive will be strengthened if amount of punishment and frustration will be increased. 3) Acceptance of dependency will be related to childen's sensitivity to parent's inhibition. Compliance to parents will be great if parents accept dependency of child. Subjects: Thirty-nine children from two years three months to three years eight months old. Nineteen children are the same subjects used in Part I a year ago. The other twenty children had beenonce interviewed for another study two years ago. Method: Standard interviewing half structured was used. One investigator became the interviewer and another the recorder. Forty-nine items on punitiveness of parent, acceptance of dependency, dependency of children and compliance of children were evaluated on five point scale. Agreement of reevaluation was 64. 5% but was raised to 97. 3% when one point disagreement was allowed. Result: 1) Correlation coefficients by Spearman among punitiveness, acceptance of dependency, dependency and compliance are shown in Table 4. 1. Main findings are as follows: (1) Punitiveness and acceptance of dependency are negatively related (r=-. 69) at 1 % level of significance.(2) Compliance is negatively related to punitiveness (r =-. 65) at 1 % level of significance, and positively related to acceptance of dendency.(r = -. 49) at 1 % level of significant.(3) Dependency seems to have no relationship to parental factors, neither punitiveness nor acceptance of dependency.(r =-. 03 or. 12) However, there are the apparent relationshii, between parental and child's variables, as its coin binations were examined, dividing subjects int (high and low groups for each variable. More than two thirds of the subjects belong to one of the following four groups as shown in Fig. 1. Group I High punitiveness-Low acceptance-Low dependency-Low compliance (N=9) Group II High punitiveness-Low acceptance-Higl dependency-Low compliance (N=5) Group III Low punitiveness-High acceptance-High dependency-High compliance (N=8) Group 17 Low punitiveness-High acceptance-Low dependency-High compliance (N=5) Thus, hypothesis I is verified by Group hypothesis 2 by Group II, and hypothesis 3 by all groups respectively. 2) Relationship between mothering at infancy an, dependency was examined. The rzcent reports o twenty mothers made from their memory abou mothering given to their children,