This experiment sought to determine the effect of various pause-deletion schemes on the comprehension of oral language under time-compression conditions. Three pause-deleted versions of a 1540-word spoken message read at 164 words per minute (wpm) were prepared. The first version deleted all inter-lexical pauses occurring at significant Immediate Constituent boundaries. The second version deleted inter-lexical pauses corresponding to Deep Structure Analogue breaks. The third deleted all inter-lexical pauses of 50 milliseconds or greater duration. These three recordings, plus a control condition with pauses intact, were differentially time-compressed to six target rates from 225 to 350 wpm. A total of 168 college students served as subjects. The subjects listened to one of the 24 experimental conditions and subsequently took a 55-item comprehension test. Results of this experiment and an additional replication of the experiment involving 192 college students failed to confirm any significant differences between pause and non-pause-deleted conditions. These results were interpreted as disconfirming a two-stage model of speech perception and comprehension. (Author/WR)