Two experiments were conducted to determine the effect of a prior stimulus (prime) on reaction time to a target stimulus. The target was a red or green color patch to which subjects responded by pressing a red or green key. The prime was a red, green, or white color patch for one group and the word Red, Green, or White for another group. The prime/target relationship was either congruent (same color), incongruent (alternate color), or neutral (white prime) and varied from trial to trial so the prime provided no information regarding the identity of the subsequent target. The interval between onset of the prime and onset of the target was 2.5 s. On one block of trials, subjects pressed the key that corresponded to the target color; on another block they pressed the noncorresponding key. In experiment 1, where the task did not require subjects to identify the prime, there was no effect of prime/target congruence. However, in experiment 2, where a secondary task required identification of the prime, reactions were faster when prime and target were congruent than when they were incongruent. This congruence effect was significant for both color and word prime groups performing under corresponding and noncorresponding mapping instructions. Results suggested that the congruence effect: (1) required processing the prime beyond the level of stimulus detection; (2) did not require a prime/target identity match; and (3) involved the stimulus encoding stage rather than the response selection stage. Findings also suggested a close interrelation between the cognitive representations of colors and color words.