TWO CLOTH SAMPLES CONTAINING STAINS WERE SENT TO 235 FORENSIC LABORATORIES, TOGETHER WITH INSTRUCTIONS TO ANALYZE THE SAMPLES AS THOUGH THEY WERE EVIDENCE COLLECTED IN CONNECTION WITH A RAPE CASE. THE LABORATORIES WERE TO IDENTIFY THE STAINS AND INDICATE THE METHODS OF ANALYSIS EMPLOYED. IN REALITY, SAMPLE A CONTAINED A SALIVA STAIN, AND SAMPLE B CONTAINED A SEMINAL STAIN. ALL THREE REFEREE LABORATORIES DETERMINED CONCLUSIVELY THAT SAMPLE B WAS SEMEN; TWO WERE INCONCLUSIVE ABOUT SAMPLE A; AND ONE DETERMINED THAT A WAS SALIVA. OF THE 128 RESPONDING TEST LABORATORIES, 47 (36.7 PERCENT) TENTATIVELY IDENTIFIED A AS SALIVA; 37 (28.9 PERCENT) WERE INCONCLUSIVE ABOUT A; 23 (18 PERCENT) CONCLUSIVELY IDENTIFIED A AS SALIVA; 16 (12.5 PERCENT) IDENTIFIED A AS NONSEMINAL; 107 (83.6 PERCENT) CONCLUSIVELY IDENTIFIED B AS SEMINAL; 18 (14.1 PERCENT) TENTATIVELY IDENTIFIED B AS SEMINAL; AND 2 WERE INCONCLUSIVE ABOUT B. THREE OF THE LABORATORIES CONCLUSIVELY OR TENTATIVELY IDENTIFIED THE SALIVA STAIN AS VAGINAL. DETAILS OF THE TEST AND REFEREE LABORATORIES' FINDINGS AND METHODS AND OF THE SAMPLE SUPPLIER'S STATEMENT ARE PRESENTED IN FIVE TABLES, INCLUDING BREAKDOWNS OF RESPONSES FROM INDIVIDUAL LABORATORIES. AN EXPLANATION OF SURVEY METHODS AND A COPY OF THE QUESTIONNAIRE ARE INCLUDED. OTHER LABORATORY PROFICIENCY TESTS ARE REPORTED IN NCJ 47522-47533 AND 47535-47542. (LKM)