Rat received 12 or 1 food pellets in alternatives discriminable by brightness, by position, or by combined brightness and position. Asymptotic running speed to the larger reinforcer and change in speeds following a reduction in reinforcer magnitude was greater with the latter two tasks, but the simultaneous contrast effect and choice behavior were relatively unaffected. Relative performance for alternative reinforcers appears to be somewhat dependent upon generalization between discriminanda, as previously suggested by others, regardless of concurrent choice behavior.After rats apparently have discriminated two distinctive runways associated with different reinforcer magnitudes, there is a tendency toward slower running speed for the larger reinforcer the smaller the reinforcer in the alternative runway (e.g., MacKinnon, 1967). However, analogous experiments employing a T-maze positiondiscrimination task have found performance for the larger reinforcer to be uninfluenced by the alternative reinforcer (e.g., Spear & Hill, 1965) or a trend opposite that found with alternative runways (e.g., Spear & Pavlik, 1966; Spear & Spitzner, 1966). Unfortunately, these relationships are not completely generalizable, since at least one experiment employing distinctive runways (Goldstein & Spence, 1963) has failed to find the effect reported by MacKinnon and others, and at least one T-maze experiment has found the effect (Spear & Spitzner, 1969). MacKinnon (1967) has interpreted this discrepancy in terms of differential tendencies for inhibition to generalize from the less favorable alternative (LFA) to the more favorable alternative (MFA). But, some of the position-discrimination mazes not yielding the effect have had equal-