A PICTURE INTEREST INVENTORY FOR THE DEAF WAS CONSTRUCTED AND STANDARDIZED. THE INVENTORY CONSISTED OF 27 TRIADS OF LINE DRAWINGS REPRESENTING VARIOUS OCCUPATIONS. ITEMS WERE GROUPED INTO 10 SCALES--PERSUASIVE, CLERICAL, MECHANICAL, SCIENTIFIC, OUTDOOR, LITERARY, COMPUTATIONAL, ARTISTIC, SOCIAL SERVICE, AND DRAMATIC. SUBJECTS WERE DEAF MALES AGED 16 OR OVER AND WITH IQ SCORES ABOVE 85 DRAWN FROM RESIDENTIAL SCHOOLS FOR THE DEAF, BUREAUS OF VOCATIONAL REHABILITATION, PUBLIC SCHOOLS, GALLAUDET COLLEGE, AND THE EMPLOYED DEAF. AFTER 3 1/2 MONTHS, TESTS WERE READMINISTERED TO 1659 OF THE ORIGINAL 2590 SUBJECTS. RAW SCORES WERE CONVERTED TO STANDARD SCORES. INSPECTION SHOWED STABILITY OF MEANS ACROSS THE RANGE OF AGES INVOLVED. TEST-RETEST RELIABILITY COEFFICIENTS FOR FOUR OF THE SUBJECT GROUPS (EXCEPTING THE EMPLOYED) FOR EACH OF THE 10 SCALES WERE MADE. MEDIAN RELIABILITY COEFFICIENTS WERE .571 FOR RESIDENTIAL SCHOOLS, .602 FOR PUBLIC SCHOOLS, .649 FOR BUREAUS OF VOCATIONAL REHABILITATION, AND .733 FOR GALLAUDET COLLEGE. TO DETERMINE VALIDITY, SCALE SCORES OF THE GALLAUDET SAMPLE WERE CORRELATED WITH CORRESPONDING SCALES OF KUDER PREFERENCE RECORD. THE PRODUCT MOMENT COEFFICIENTS RANGED FROM .259 TO .612 WITH A MEDIAN OF .417. AS ANOTHER MEASURE OF VALIDITY, THE TEST WAS ADMINISTERED TO A SAMPLE OF EMPLOYED DEAF IN 22 OCCUPATIONS. MEANS AND STANDARD DEVIATIONS ON EACH SCALE FOR THE 22 OCCUPATIONS WERE DERIVED. ALTHOUGH THE SCALES DID TEND TO DISCRIMINATE AMONG THE OCCUPATIONAL GROUPS, SOME SCALES DID NOT DISCRIMINATE IN THE EXPECTED DIRECTION. INTERCORRELATIONS WERE COMPUTED AMONG THE 10 SCALES ON THE SAMPLE GROUPS. A CONSIDERABLE AMOUNT OF OVERLAP IN WHAT IS MEASURED BY EACH SCALE WAS NOTED. RELIABILITY OF THE DIFFERENCE BETWEEN SEPARATE SCALES YIELDS A RANGE OF .00 TO .79 WITH AN AVERAGE OF .56. NORMS IN T SCORE TERMS ARE GIVEN FOR ALL GRADES IN THE RESIDENTIAL AND PUBLIC SCHOOLS AND FOR 15 OCCUPATIONAL GROUPS. TO DETERMINE THE CORRECTNESS OF PLACING ITEMS IN SCALES, A BISERIAL CORRELATION BETWEEN EACH ITEM ALTERNATIVE AND EACH OF THE 10 SCALE SCORES WAS COMPUTED. NECESSARY STEPS TO MAKE THE TEST READY FOR OPERATIONAL USE ARE DESCRIBED. (FL)