Comparative advertising is widely used as a persuasion tool. Contemporary advertisers employ both direct (e.g. "Total Corn Flakes have more nutritional ingredients than Kellogg Corn Flakes") and indirect comparative advertisements (e.g. "Total Corn Flakes have more nutritional ingredients than other corn flakes") in their campaigns. Consumers process ads predominantly either through analytical (using reason and semantics) or imagery (using nonverbal, sensory representation of perceptual information) modes of processing. Drawing on extant research in advertising and information processing, we posit that direct comparisons are better suited for analytical processing while indirect ones are more amenable for imagery processing. Using a 2*2 factorial design (comparison format-direct/indirect; mode of information processing-analytical/imagery), we show that direct (indirect) comparative advertisements are more effective in reducing perceived manipulative intent, enhancing attitude toward the advertisement, and increasing the perceived differences between the brands for consumers using analytical (imagery) information processing modes. We add to the comparative advertising literature by identifying a moderator (mode of information processing) that influences the effectiveness of direct and indirect comparisons. Further, we study indirect comparisons when most extant research considers direct comparisons. Managerially, our work suggests that advertisers using comparative advertising should use direct comparisons under analytical processing conditions and use indirect comparisons under imagery processing conditions. © 2013 John Wiley & Sons, Ltd.