Researchers have long been aware that intention to vote, reported vote and validated vote are different. In particular, a perfect validation would be the gold standard, but doing validation well is time-consuming and expensive. This chapter focuses on the best American data for comparing the three measures of individual turnout. It also focuses on American National Elections Studies (ANES) data from 1980, 1984 and 1988, the only years in which all the relevant variables were available. The chapter discusses limitation of the ANES data. It examines whether the three dependent variables such as intention to vote, reported vote and validated vote have the same relation to the customary explanatory variables. The sample is those who intended to vote, excluding those who intended to abstain, nearly all of whom are non-voters. The table assesses whether carrying out the intention to vote is related to socio-demographic characteristics and political attitudes.