The purpose of this paper is to lay a theoretical foundation for the revision of the Test of Spoken English. The revision project was undertaken in response both to concerns expressed by researchers and score users about the validity of the TSE test and to a request by the TOEFL Committee of Examiners to make the TSE test more reflective of current thinking on the assessment of oral language skills. In the paper, we first discuss communicative competence as a basis for understanding the nature of language knowledge, followed by discussions of sociolinguistic and discourse factors which influence spoken language performance. Test method characteristics which influence test performance are also discussed, as are types of evidence necessary for establishing reliability and validity of the revised TSE test. The paper concludes with a discussion of the implications of the theory for the interpretation of examinee performance with regard to academic and professional contexts of language use.