AN AUDIOVISUAL FILM WAS PRODUCED TO HELP RESEARCHERS AND CLINICIANS ESTABLISH THEIR RELIABILITY OF JUDGMENTS OF MISARTICULATIONS IN CHILDREN'S SPEECH. THE FINAL FILM SCRIPT WAS PREPARED AFTER A PRELIMINARY STUDY HAD BEEN EVALUATED BY PANELS OF BOTH TRAINED AND UNTRAINED LISTENERS. SIX CHILDREN, BETWEEN THE AGES OF 6 AND 13, WHOSE MISARTICULATION RANGED FROM MODERATE TO SEVERE RECORDED 23 WORDS (CONTAINING 43 CONSONANT SOUNDS) AS SINGLE WORDS, PHRASES AND TRIOS IN RANDOM ORDER TO GIVE A TOTAL OF 558 SOUNDS TO BE JUDGED FOR ACCURACY. TWO FORMS OF THE AUDIOVISUAL FILM WERE PRODUCED AND WERE EVALUATED BY THREE GROUPS OF JUDGES--50 CLINICIANS, 50 SENIORS, MAJORING IN SPEECH AND HEARING, AND 50 CLASSROOM TEACHERS. STATISTICAL EVALUATION OF THE DATA INDICATED THAT THE AUDIOVISUAL TEST WAS VALID, RELIABLE, AND EASY TO ADMINISTER. CLINICIANS AND GRADUATING SENIORS LARGELY AGREED BUT DIFFERED FROM TEACHERS IN IDENTIFYING MISARTICULATIONS. IDENTIFICATION OF MISARTICULATIONS IN ISOLATED WORDS WAS BETTER THAN IN PHRASES FOR JUDGES IN ALL GROUPS. THE AUDIOVISUAL METHOD WAS SUPERIOR TO AN AUDIO-ONLY METHOD FOR ALL GROUPS. (JC)