Although several studies have reported relaxing and stimulating effects of odors on physiology and behavior, little is known about their underlying mechanisms. It has been proposed that participant expectancy could explain these activation effects. Since emotional stimuli are known to modulate time perception, here we used the temporal bisection task to determine whether odors have objective relaxing and stimulating effects by respectively slowing down or speeding up the internal clock and whether prior expectancy could alter these effects. In Experiment 1, 118 participants were presented either with a strawberry odor or an odorless blank. In Experiment 2, 132 participants were presented either with a lemon odor or an odorless blank. In both experiments, expectancy was manipulated using suggestion (verbal instructions). The stimulus was either described as relaxing or stimulating, or was not described. In the absence of prior suggestion, findings showed that, compared to participants presented with an odorless blank, participants presented with the strawberry odor underestimated sound durations (i.e., a relaxing effect) whereas participants presented with the lemon odor overestimated them (i.e., a stimulating effect). These results confirm that pleasant odors can have objective relaxing and stimulating effects by themselves, which are better explained by arousal-based mechanisms rather than attentional distraction. Furthermore, in both experiments, incongruent suggestions undid the effects of both odors without reversing them completely (i.e., strawberry did not become stimulating even if participants were told so). Both these bottom-up and top-down influences should be considered when investigating the emotional impact of odors on human behavior.