Different intertrial intervals (lTIs) may influence the aversiveness of the goal area in aversive training with one-way procedures. That possibility was explored. Albino rats were given 20 shock-escape runway trials with short or long ITIs followed by 5 extinction test trials with goalbox measures for latency to eat, climb out, and move onto a wood floor. Different measures produced different outcomes. "Time to eat" did not differentiate groups. Massed-trials animals were quicker to "move over" initially but slowed down over trials. Spaced and massed procedures did not produce initial differences on "climb out," but spaced-trials animals speeded up over trials and massedtrials animals slowed down. Results suggested that differences in ITIs did not generate major differences in levels of aversion in the "safe" area. ITI effects on performance may be due to differential effects on the conditioning process.Aversive training in one-way apparatuses versus single compartment and two-way apparatuses pose different problems with respect to the relative aversiveness of the conditioned stimulus (CS) context, and post-unconditioned stimulus (US) conditions. In a one-way apparatus, such as the straight runway, a typical procedure is to apply the CS and US in the context of the startbox and alley, follow this with a period of time in the "safe" goalbox, and then have the animal spend the intertrial interval (lTI) in a retaining cage dissimilar from the runway apparatus. Except for a short period in the startbox at the beginning of each trial, the US is consistently present with the CS, startbox, and alley, whereas the US is consistently absent during the goalbox and IT! confmement periods. In contrast, a typical procedure in a two-way shuttlebox is to present the CS and shock in one compartment, allow escape to an adjoining compartment for a period of time during which the CS and US are absent, and then reinitiate the CS and US in that same compartment, with subsequent escape back to the initial compartment. In this two-way procedure, not only is the "safe" period of confinement the same as the ITI, but the two compartments are altematingly aversive and safe.Despite the absence of the US in the goalbox and ITI periods in the one-way procedure, an argument can be made that those periods are not actually devoid of aversive influence. If that is true, the presence of aversiveness in the goalbox could complicate interpretations of self-punitive responding (Babb, Kostyla, & Bennett, 1980;Brown, 1969;Dean & Pittman, 1991) but might help to elucidate findings in which small frequencies of goal punishment, with short ITIs, produce facilitated responding and longer frequencies produce suppression (Matthews