People often encounter large bodies of information with environmental significance (e.g., data on emissions from flights). To evaluate the environmental footprint of such data, it must be integrated, a process that can be biased depending on stimulus characteristics. In two experiments, the present study showed that carbon dioxide emissions are perceived as more environmentally harmful when drawn from a negatively skewed stimulus distribution and more environmentally friendly when drawn from a positively skewed distribution, even when the sum of the emissions is identical across the two distributions. This distribution-density effect was particularly strong when participants had context foreknowledge about the stimulus distribution’s range and shape. A third experiment found that real-world familiarity with a stimulus distribution does not provide the same effect as this type of context foreknowledge. Stimulus structures influence how people evaluate environmental impact, and more knowledge can sometimes increase rather than decrease susceptibility to cognitive bias.