Cognitive biases are heuristics that shape individual preferences and decisions in a way that is at odds with means-end rationality. The effects of cognitive biases on governing are underexplored. The authors study how election administrators’ cognitive biases shape their preferences for e-voting technology using data from a national survey of local election officials. The technology acceptance model, which employs a rational, means-end perspective, suggests that the perceived benefits of e-voting machines explain their popularity. But findings indicate that cognitive biases also play a role, even after controlling for the perceived benefits and costs of the technology. The findings point to a novel cognitive bias that is of particular interest to research on e-government: officials who have a general faith in technology are attracted to more innovative alternatives. The authors also find that local election officials who prefer e-voting machines do so in part because they overvalue the technology they already possess and because they are overly confident in their own judgment.