Recent forms of the new three-section Test of English as a Foreign Language (TOEFL) were administered to 88 native speakers of English just prior to graduation from a college-preparatory high school program. Total test score distributions were highly negatively skewed, reinforcing findings of earlier studies that the TOEFL is not psychometrically appropriate for discriminating among native speakers of English with respect to English language competence. Although the native English examinees achieved high total test scores and performed excellently on the listening comprehension section, a number of items in the other two sections (structure and written expression; reading comprehension and vocabulary) were answered incorrectly by over 20 percent of the examinee group. Included in these sections were a number of questions considered by the TOEFL test development staff as dealing with basic grammar points or straightforward reading passages of a type that college-level students should be expected to handle without difficulty. From these results, it is concluded that although response data from native English speaking examinees may be of some use in designating particular test questions for closer examination, errors made by college-bound native speakers should not automatically be considered indicative of item inappropriateness for the TOEFL population. (Author)