Auditory streaming and its relevance to attentional processing was examined using event-related brain potentials (ERPs) in situations facilitating perception of one or two streams of sounds. Subjects listened to sequences of brief tones of three different frequencies presented in random order. In evenly spaced (ES) conditions, the three frequencies were equidistant on the musical scale. In clustered, easy (CE) conditions, the attended frequency was distinct, while the middle and extreme distractor tones were clustered together. In clustered, hard (CH) conditions, the attended frequency was clustered with one of the distractors. The subjects pressed a button in response to occasional target tones of longer duration at a prespecified frequency. The subjects were faster and more accurate in CE conditions than they were in ES conditions, and ERP attention effects were enhanced in amplitude in CE conditions. Conversely, the subjects were slower and less accurate in CH conditions and ERP attention effects were delayed in latency and decreased in amplitude. Clustering effects suggest that the processing of stimuli belonging to the attended stream was promoted and the processing of those falling outside the stream was inhibited. The timing and scalp distribution of clustering-related changes in ERPs suggest that clustering modulates early sensory processing in auditory cortex.The clustering by frequency of successive acoustic events influences our ability to distinguish auditory stimuli from one another. For example, order judgments of target tones embedded in short, frequently repeated sequences are improved when distractor tones are clustered together (Bregman & Rudnicky, 1975;Jones, Kidd, & Wetzel, 1981). The clustering ofdistractors has also been shown to improve the speed as well as the accuracy of target detection, even when the tones are presented in random order . Conversely, the clustering of attended tones with distractors degrades performance and slows reaction times (RTs; Alain, Achim, & Richer, 1993). These performance variations can be explained by the clustering-based segregation of tone sequences into distinct perceptual streams. Improved discriminability is observed when the distractors are segregated in a perceptual stream distinct from the targets. On the other hand, discriminability is decreased when targets and distractors are grouped together. Bregman (1990) proposed that auditory stream segregation is the result of preattentive perceptual grouping This research was supported by grants from the NIDCD and the VA Research Service to D.LW., and by a postdoctoral fellowship provided by the Fonds de la Recherche en Sante du Quebec to C.A. Our special thanks to Greg Shenaut and John Lackey for software development. We would also like to thank K. H. Ogawa, K. Alho, R. Naatanen, M. G. Woldorff, A. S. Bregman, L. Krueger, and an anonymous reviewer for helpful comments. Correspondence should be addressed to Claude Alain, Department ofNeurology, Northern California Systems of Clinics (612/127), ISO Muir Road...