The Consultation Judgment Inventory is a paper-and-pencil instrument designed in 1983 to measure ability to make appropriate clinical judgments regarding possible actions to take in psychological consultation situations (4). It was revised (3) to improve certain psychometric characteristics. Each of the 33 items consists of a brief consultant-consultee dialogue in which the consultee presents a commonly occurring consultation situation. Five multiple-choice answers representing possible actions were provided for each situation, with one being keyed as correct. Commonly occurring situations and actions keyed as correct were based upon those recommended (1, 2). Overall scores on the instrument can range from 0 to 33. The 1985 revision was administered to 62 undergraduate psychology majors and 41 graduate predoctoral counseling psychology majors; none had ever received any training in consultation. The undergraduate sample consisted of 30 white women, 18 black women, 13 white men, and 1 black man; the mean age of the undergraduates was 21.1 yr. The graduate sample consisted of 23 white women, 15 white men, and 3 black women; the mean age of the graduate sample was 24.5 yr. Undergraduate subjects were used solely to provide a basis for assessing whether some graduate training, even though not specifically in consultation, improves performance. The mean over-all inventory score for graduate students was 16.29, compared to 9.01 for undergraduates. The remainder of the analysis involved the 4 1 graduate students. The Kuder-Richardson 20 reliability coefficient obtained for this group was .87. Item difficulties ranged from ,195 to ,829. Discrimination indices ranged from -.090 to .908. One item was seen as too difficult (with a difficulty index of ,195) and five items too easy or answerable based upon "generic" professional knowledge (with item difficulty indexes of ,732, ,732, ,780, ,805, and ,829, respectively). Three items were seen as not discriminating sufficiently between high and low scores; discrimination indexes were -.09, .09, and .09, respectively. Further work is planned to improve item difficulties and discrimination indexes.