The stability of House Drawing in H-T-P test was examined by administering the test to 38 high school male students (Group A), 40 high school male students (Group B) and 32 juvenile delinquents (Group C). Group A performed the test two times, two weeks apart. The two sets of drawings were then statistically analysed by the experimenter. And the matching of 20 drawings of the first time with 20 drawings of the second was attempted by four judges. In result, test-retest correlations were significant in the style of drawings and the impression as a whole but not so significant in details. Group B and Group C also performed the test two times but the instruction was changed in the second time. Result showed that the stability of drawings was maintained in Group C but not in Group B. Line pressure, One wall and Window Curtains were stable in drawings of all groups.
