We examined whether different types of brain images affect readers' evaluations of scientific reports. Five different brain images were selected from the neuroscience literature: a whole brain, an inflated brain, a cross-sectional brain slice, a glass brain, and a topographic map. First, the images were subjectively rated by 31 nonexperts for qualities associated with realism and perceived complexity. Each of the five images was later presented alongside one of five fictitious neuroscience articles (image-text pairings counterbalanced), and a different group of 122 novices rated the accompanying articles for scientific reasoning. They also separately reported their familiarity with each image type. Brain images previously rated as more three-dimensional produced more positive evaluations of the articles with which they were presented. Perceived image complexity also showed a marginal nonlinear relationship with article credibility ratings. Our findings suggest that choice of image format matters when disseminating neuroscience research to the general public.