Theories of voter turnout have focused almost exclusively on the costs and benefits of voting, even though one potentially important aspect of turnout is what takes place after a citizen has decided she intends to vote but before she arrives at the polls. I test two social psychology theories affecting this process: self-prophecy and implementation intentions. The self-prophecy effect occurs when an individual is asked to predict whether he will vote, and this act of prediction alters his future behavior by making him more likely to vote. Applying the implementation intentions theory to voting, citizens would be more likely to vote if they focused on the concrete actions involved in getting to the polling location rather than simply on the end-state of having voted. The field experiment consisted of phone calls to 5,609 registered voters with scripts intended to test these theories. Neither treatment generates a statistically significant result, which undermines the reliability of very large effects found in prior research on voting self-prophecy effects, and suggests the impact of implementation intentions may be limited with respect to voting.