There is sufficient evidence from both L1 and L2 research to in dicate that temporal variables (including speech rate and pause pheno mena) can be manipulated to facilitate listening comprehension. There is, however, no reliable evidence to indicate that rate modification is a widely occurring phenomena in NS-NNS interaction. This controlled study was designed to investigate whether 6 university professors would modify the temporal organization of their speech when deli vering identical lectures to NS, low-proficiency and high-proficiency NNS groups. Samples of the recorded lectures were spectrographi cally analysed and a large number of temporal variables were investi gated. Contrary to expectations derived from a reading of the L2/ESL literature no significant modifications to speech rate were found in deliveries to either the low- or high proficiency groups. Observed changes in articulation rate narrowly missed being significant, but differences were between the two NNS groups, not between the NS and NNS groups. Highly significant differences were, however, observed in the data on filled pauses. L1 pausological literature is drawn upon in suggesting explanations of these findings. Implica tions for L2 teaching are discussed.