Contemporary explanations of the trial-spacing effect (TSE) were evaluated. Experiment 1 revealed, in rats given tone-footshock trials with 15-, 60-, or 900-s intertrial intervals (ITIs), a direct relationship between freezing to the tone and ITI (the TSE) but an inverted U-shaped relationship between freezing to the training context and ITI. In Experiment 2, footshock preexposure eliminated the TSE that otherwise occurs across 15- to 60-s ITIs but had no effect on the TSE that occurs across 60- to 900-s ITIs. In Experiments 3 and 4, neither (a) increasing posttraining exposure to the training context in rats trained with 60-s ITIs nor (b) reducing between-trial exposure to this context in rats trained with 900-s ITIs influenced freezing to the tone. These findings suggest that the TSE obtained in this research is due to more than 1 mechanism: 1 responsible for the TSE that occurs with ITIs less than approximately 60 s and another responsible for the TSE that occurs with ITIs greater than this. Although the perceptual-defensive-recuperative model may correctly describe the former mechanism, none of the theories tested seems to correctly describe the latter.