Images are an integral part of any visual advertisement, regardless of medium. Whether online banner, digital screen, or wall poster, the decisions of professional designers on how to compose those images are often based on specific rules of composition that describe proper cropping or object placement techniques. However, existing research shows that even individual visual elements can affect the complexity of an image and influence cognitive processing, which, in turn, can affect not only attention or memory, but also the emotional reaction of the viewer. This study aims to explore the influences of some popular composition techniques on the perceived complexity of an image, and their possible impacts on the effectiveness of visual advertisements – specifically, public service announcements (PSAs) promoting pro-environmental behavior. To this end, two experiments were conducted. The first experiment explored the role of image composition characteristics on the perceived complexity of a picture, and evoked emotions by manipulating the level of cropping and object merging in images. The results of this experiment demonstrated that composition affected image complexity ratings, and at the same time that simplicity correlated with the positive valence of the emotional reaction evoked. The second experiment assessed the impact of image complexity on the effectiveness of PSAs. After observing PSAs containing more complex images, participants demonstrated significantly higher intentions to behave pro-environmentally. This effect was not influenced by the different types of text-based messages used in the PSAs. These findings demonstrate that composition is an important factor that can have a significant impact on the effectiveness of visual advertisements by influencing the attention, emotions, and intentions of viewers of advertisements.