What is the role of sensory experience in the acquisition of knowledge? Locke famously argued that a person born blind could never understand anything about color. Contrary to this prediction, prior studies have found that congenitally blind individuals do acquire some information about color. But questions remain about the depth and nature of this knowledge. Do blind individuals merely learn verbal associations (e.g., that bananas are yellow)? Alternatively, could blind people’s color knowledge be as structured, coherent, and inferentially rich as that of sighted people? In a series of experiments, we examined blind and sighted people’s knowledge about the colors of natural and manmade objects. Relative to the sighted, blind individuals were less consistent in their knowledge about the common colors of specific objects (e.g., banana-yellow). Blind and sighted participants, however, made identical inferences about how likely two instances of the same object kind are likely to have the same color, for familiar and novel objects. When asked to explain their own knowledge about the origins of object color, blind individuals also gave detailed and sophisticated explanations, just like the sighted. Based on these data, I will argue that causal theories about “sensory” categories are acquired even in the absence of first-person sensory experience.