The persistence of attention-deficit/hyperactivity disorder (ADHD) into adulthood, particularly in women, is an unresolved question. Using a diverse, prospectively followed female sample (ADHD: n = 140; comparison: n = 88), we examined (a) informant (parent- vs. self-report) and (b) symptom threshold (DSM-IV cutoffs vs. a developmentally referenced criterion [DRC]) with respect to estimates of ADHD persistence into young adulthood (M = 19.6 years). We also ascertained the linkages of ADHD status (as measured by each informant and via the different symptom cutoffs) with impairment in key outcome domains. Informant diagnostic concordance was significant but small (κ = 0.22). Via parent report, more childhood-ascertained probands (44 %) than comparison participants (1 %) met full ADHD criteria in adulthood (p < 0.001, ϕ = 0.47); parallel results were found via self-report (22 % vs. 2 %; p < 0.001, ϕ = 0.28). As expected, DRC designations yielded higher persistence estimates than DSM cutoffs. With control of key covariates, parent-reported ADHD was independently associated with poor outcomes on eight of nine measures but self-reported ADHD was associated only with low math scores. Also, participants who met the DRC cutoff but not the higher DSM threshold showed similar impairment to DSM-diagnosed participants, far greater than that of participants who met neither cutoff. A lowered symptom threshold in adulthood compared to childhood appears justified.