Abstract Aims The purpose of this study is to determine which respondents of national surveys who report low ‘usual’ past year drinking frequency, are unsure of their drinking frequency or refuse to answer, are misclassified either as ‘current drinkers’ or as ‘ex-drinkers.’ Design and setting The data are from the 2000 National Alcohol Survey, a national household telephone probability sample of adults in all 50 US States and Washington, DC (n = 7612). Participants A subsample of 1734 respondents who reported drinking ‘less than once a month but at least once a year’, ‘less than once a year’, ‘refused’ and ‘don’t knows’ were also asked if they had consumed a whole drink of any alcoholic beverage in the 12 months using dates to anchor the question. Measurements Five alcohol measures were used to compare groups who changed or did not change their drinking status based on the followup question: volume, mean number of days 5 +, social consequences, dependence symptoms and DSM-IV. Findings The results indicate that only 4.6% of respondents (n = 349) changed their drinking status, and the percentages were equal in both directions. Women were more likely to change their drinking status and specifically become ex-drinkers based on the follow-up question. Respondents who became current drinkers drank significantly at a lower average volume than those who remained current drinkers. The prevalence rates based on the five alcohol measures were not affected by how ‘current drinker’ is defined. Conclusions Adding the follow-up question is recommended for new alcohol surveys so that fewer respondents are misclassified.