Studies on sine-wave speech (SWS) perception suggest that formants contain sufficient information for sentence intelligibility. This study further investigated the effects of amplitude modulation, number of sine-waves, and vowel resonance in SWS recognition. Results showed that Mandarin sentences synthesized using frequency trajectories of the first two formants were highly intelligible with additional contributions from formant amplitude modulation. However, amplitude modulation significantly contributed to intelligibility when only the vowels were preserved. The present work demonstrates that the intelligibility of Mandarin SWS can be largely attributed to the frequency transition of the first two formants and is susceptible to temporal interruption.