Detailed data were collected on lifetime caffeine intake from 202 Vermont residents using a random-digit dial telephone survey. The sample appeared representative and test-retest reliability of caffeine intake was high (r = .95). Almost all participants (96%) had ever used and most (83%) presently used one or more caffeinated beverages weekly. The average caffeine intake was 186 mg/day. Many caffeine users (61%) used caffeinated beverages other than coffee. Current caffeine intake was a poor measure of lifetime intake. For example, among ever-users of caffeine, 41% had stopped at least 1 type of caffeinated beverage and 14% had stopped caffeine altogether. Cessation was mostly due to health concerns and unpleasant side effects. It was concluded that simply asking about "usual" coffee use is a poor and biased estimate of lifetime caffeine use. Thus, prior findings that caffeine is not associated with medical or behavioral disorders may represent false-negative results.