Participants named (Experiments 1-2) or categorized (Experiments 3-4) the picture or the word of picture-word compounds that varied in the relative saliency of the 2 components and in the correlation between them over the experimental trials. Picture-word interference (PWI) was gauged through Stroop and Garner effects. PWI was found to be malleable; its magnitude and very presence depending lawfully on the contextual variations introduced. The contingent nature of PWI is a fact to be reckoned with by theorists of picture-word processing.