The purpose of this study was to attempt to determine if score gains obtained upon repeated testing with an intelligence test result from a practice effect, from students remembering specific items, or from a combination of both. The verbal and nonverbal batteries of an I.Q. test were administered to 860 sixth graders on three occasions with two-month and four-month intervals between testing sessions. Some students received the same form of the test each time they were tested while others received alternate forms of the test. The results indicated that the subjects did experience an increase in verbal mean I.Q. In the nonverbal results, only groups retested with the same form of the test experienced significant mean gains. The verbal mean gains appeared to result from a practice effect while the nonverbal results appeared to result from students remembering specific items from one testing session to the next. (Author)