In an experiment on the effect of intensity accents on the perception of time intervals between tones, H. G. Tekman (2001) found that the regular placement of deviant time intervals in short sequences of tones reduced detection, especially if intensity accents marked the deviant time intervals. That was the opposite of what one would have expected on the basis of the dynamic attending theory of M. R. Jones (1976). The effect might have occurred because temporally deviant tones create cumulative onset shifts that affect all the subsequent tones. If the deviations were randomly placed, then they could follow each other in close succession and change the local tempo. In the present study, the changes of local tempo, which might have acted as a cue for the detection of temporal deviations in the random sequences, were eliminated by compensating for deviant time intervals with equal deviations in the opposite direction in the interval that followed. That change in the stimuli eliminated the negative main effect of regularity, and the accenting interacted with regularity in favor of detection in the regular sequences. However, a simple advantage of regular over random sequences was not observed. The author discusses possible reasons for the lack of a facilitatory effect of regularity.