There is a long debate on the relative importance of spectral and temporal cues in speech perception theories. On the one hand, the highly-intelligible sine-wave speech (SWS) has been viewed as a representation of the global spectral structure of the speech signal. On the other hand, there is accumulating evidence showing that the temporal aspects of speech without spectral details provide sufficient speech understanding. The present study explored whether the temporal envelopes imbedded in the SWS contribute to its intelligibility. In the experiments, both SWS and natural speech signals were processed with noise and tone vocoders to remove the spectral details but to preserve the temporal envelopes. Twenty-two normal-hearing, native English-speaking adult listeners participated in sentence recognition tasks. Speech recognition performance of vocoder-processed SWS was slightly inferior to that of vocoder-processed natural speech but both reached plateau performance at 6–8 channels. Acoustic analysis further indicated that the temporal envelopes of the SWS were almost identical to those of the natural speech, with a mean correlation coefficient r = 0.949 across all sentences. The results provide strong evidence that the SWS represents both spectral and temporal structures of the speech and that the temporal envelopes imbedded in SWS carry important information for speech recognition.