This discussion paper attempts to take a comprehensive look at electoral integrity. The concept is initially defined in a way which follows the ordinary language use of “integrity”, i.e. as something which has not in any way been damaged or ruined. Electoral integrity thus refers to the perfect election or at least an election where no one has anything to complain about. In real life, however, such elections are yet to come, but the profession lacks an across-the-board concept, which covers what is sometimes – but often somewhat vaguely or imprecisely – termed “free and fair”, “good”, or “acceptable” elections. “Electoral acceptability” is maybe the concept best covering “real-world electoral integrity”, so that the relationship between the two might remind one of the relationship between “Dahl’s democracy” and “polyarchy”? The opening discussion is followed by a presentation and discussion of some key points in relation to the eight attributes of electoral integrity suggested here. The conclusion is that electoral integrity is a complicated concept to work with and that it is not enough to measure it solely based on election observation reports or the average of perceptions of a sample of voters or so-called experts. To the extent possible, hard data should also be taken into account.