THE FIRST EXPERIMENT EXPLORED THE INTERVIEW TECHNIQUES OF FREE REPORT, THE USE OF GENERAL QUESTIONS, AND THE USE OF SPECIFIC QUESTIONS. FREE REPORT WAS EXPECTED TO PRODUCE THE MOST ACCURATE BUT LEAST COMPLETE INITIAL RECALL, WITH THE INCOMPLETENESS HAVING A DETERIMENTAL EFFECT UPON THE ACCURACY OF LATER RECALL. FREE REPORT DID PRODUCE THE INITIAL ACCURACY AND INCOMPLETENESS EXPECTED, BUT THERE WAS NO DAMAGE TO THE ACCURACY OF LATER, MORE STRUCTURED RECALL. THE SECOND EXPERIMENT INVESTIGATED THE USE OF THE FREE REPORT TECHNIQUE ALONE OVER A 2-MONTH PERIOD. THE EFFECT UPON RECALL ACCURACY OF A DELAY OF 2 WEEKS OR 2 MONTHS FROM THE TIME OF THE WITNESSED INCIDENT TO THE FIRST RECALL SESSION WAS INVESTIGATED. A HIGH LEVEL OF ACCURACY WAS ACHIEVED AT ALL RECALL SESSIONS; HOWEVER, THE DELAY HAD AN ADVERSE EFFECT UPON THE COMPLETENESS OF RECALL. USING GENERAL QUESTIONS IN AN INTERVIEW IS APPARENTLY THE OPTIMAL TECHNIQUE FOR OBTAINING EVIDENCE FROM CHILDREN, SINCE THE SUBJECTS INTERVIEWED WITH THIS TECHNIQUE HAD MORE CORRECT POINTS WITHOUT SIGNIFICANTLY MORE INCORRECT POINTS THAN THE OTHER TECHNIQUES; HOWEVER, FREE REPORT WAS THE MOST ACCURATE REGARDING PEOPLE DESCRIPTIONS. THESE FINDINGS SUPPORT THOSE OF SIMILAR STUDIES. TABULAR DATA AND REFERENCES ARE PROVIDED. (RCB)