The effect of stimulus salience on latent inhibition (the retardation of conditioning produced by prior exposure to the event to be used as the conditioned stimulus [CS]) was examined in an experiment using rats as subjects and the conditioned suppression procedure. The stimuli were a more salient light and a less salient tone—rats trained with light as the CS showed more suppression than rats trained with the tone as the CS; and rats tested with tone and light separately after conditioning with a CS consisting of a tone + light compound showed more suppression to the light than the tone. This pattern of results was reversed, however, in subjects given a series of nonreinforced presentations of the tone and the light separately prior to conditioning with the compound. We conclude that latent inhibition develops more readily for the more salient stimulus and that its effects can outweigh those that derive from the intrinsic salience of the stimulus. Theories of latent inhibition that predict, or can accommodate, this conclusion are considered.