Because of the COVID-19 threat to in-person voting in the November 2020 election, state and local election officials have pivoted to mail-in voting as a potential solution. This method of voting—while safe from a public health standpoint—comes with its own set of problems, as increased use of mail voting risks amplifying existing discrepancies in rejected mail ballots. While some mail ballot rejections are to be expected, a lack of uniformity in whose ballots get rejected among subgroups of voters—whether for mistakes on a ballot return envelope (BRE) or lateness—raise concerns about equal representation. We draw on official statewide voter file and mail-in ballot data from the 2018 midterm election in Georgia, a state that until the pandemic did not have widespread use of mail voting, to test whether some voters are more likely to cast a mail ballot that does not count. Most importantly, we distinguish between ballots rejected for lateness and those rejected for a mistake on the return envelope. We find that newly registered, young, and minority voters have higher rejection rates compared with their counterparts.