In this study we identify some of the phonetic parameters that correlate with nonnative speakers' (NNSs) perceptual judgments of accent in English and investigate NNS listener perceptions of English from a World Englishes point of view. Our main experiment involved 3,200 assessments of the perceived degree of accent in English of two speaker groups: 11 Japanese and 5 Americans. Two additional and separate phonetically untrained listener groups, one composed of 10 Japanese and the other of 5 Americans, did the perceptual assessments. A follow-up auditory analysis by two phonetically trained listeners and an acoustic analysis showed that the untrained Japanese listeners used primarily nonsegmental parameters (intonation, fluency, and speech rate) to make perceptual judgments, whereas segmental parameters had a relatively minor role. Untrained American listeners exhibited the opposite pattern: Segmentals (especially /r/ and /l/) figured prominently, and nonsegmentals played a relatively minor role. Our study shows how native-speaking (NS) and NNS listeners perceive degree of accent in English in fundamentally different ways, each based on different phonetic parameters. We consider the implications that our findings might have for a recently proposed phonological syllabus for English as an international language (EIL) designed with NNS-NNS interactions in mind. (Contains 5 figures and 2 tables.)