A study examined procedures used to test the oral communication skills of students between the ages of 15 and 18 in the United Kingdom, with special attention given to strategies which might be of value to adopt in the United States. Subjects were 4 pairs of students tested from a randomly selected sample of 112 schools. Twenty-two assessors were recruited from schools not involved in the assessment program. They were trained first by listening to two cassette tapes containing several examples of each student task exemplifying a range of performances obtained during a pilot project. Secondly, they were given intensive training in a two-day session at National Foundation for Educational Research (NFER) offices. A test called for students to perform 7 different tasks. Performances were assessed by means of both impressionistic and analytic marking. Three general criteria were used to assess the sequential structure, lexico-grammatical features, and performance features. Results showed that standardized performance testing in speech on a large scale is possible when resources and a commitment are present. The closest corollary we have in the United States is forensics assessment. Similarities between the two programs include: both define specific speaking tasks for the student; both involve a period of preparation with a tutor or coach; and both involve evaluation by an outside evaluator. Dissimilarities include: English Speaking Board (ESB) tasks more closely adhere to real-life communication than most forensics events and involve job-related skills; and forensics events are more competitive and tend to involve only competent speakers. (RAE)