Rats judged inter-reward intervals (IRIs) in a two-alternative, forced-choice task. The IRI was either short (S) or long (L). At the end of each IRI, two response levers were inserted into the box. A press to the lever designated as correct or incorrect produced a large (3-5 food pellets) or small (1 pellet) reward, respectively. Psychophysical functions were obtained by testing intermediate IRIs, which were followed by the small reward, independent of the response location. S and L intervals were manipulated across groups, as specified (in seconds) by the group name, S-L: 3-12 (n=12), 25-100 (n=12), and 50-200 (n=7). The psychophysical functions (p[L] vs. IRI) were ogival in shape and had bisection points (p[S]=p[L]=0.5) near the geometric mean of S and L intervals. The psychophysical functions did not superimpose in relative time (IRI/L). Instead, 3-12 was timed with greater relative sensitivity than were 25-100 and 50-200.