Background Spoken discourse production has been increasingly recognised as an important source for assessing language competence of persons with aphasia (PWA). Most linguistic measures rely on transcripts of spoken discourse, so they have seldom been used in clinical practice. Thus, there is a need for standardized and norm-referenced measures that can be easily applied in clinical settings.Aims The present study aims to develop core lexicon lists for Mandarin Chinese PWA using AphasiaBank data elicited with different discourse production tasks and test their validity in detecting language deficiency.Methods & Procedures We selected seven semispontaneous discourse tasks within three major task types, i.e., picture descriptions, procedural discourse and story narrative, for core lexicon list development and for examining task type effects. Transcripts from 61 control speakers were split into two sets with age – and education matched: one set (n = 43) as a normative sample used to generate the core lexicon lists and the other (n = 18) as a healthy control group served as a baseline for comparison with PWA productions. Transcripts of 18 age- and education- matched PWA were retrieved from the AphasiaBank database. We selected the top 30 words with the widest distribution in the normative sample to generate core lexicon lists. Then for both PWA and control groups, we calculated agreement percentage as core lexicon scores.Outcomes & Results Core lexicon scores of PWA were significantly lower than the control group participants for all seven tasks and positively correlated with aphasia severity for two story narrative tasks, suggesting that story narrative tasks were more effective in assessing discourse productions compared with picture and procedural descriptions. In addition, core lexicon scores also correlated with lexical diversity (R scores) and discourse informativeness (Correct Information Units) for all seven tasks.Conclusions & Implications Core lexicon scores can effectively distinguish discourse production by Mandarin PWA and healthy controls and they reflect lexical diversity and informativeness of discourse. Our findings, thus, have clinical implications in that these lists could serve as an alternative transcription-less approach in quick Mandarin PWA discourse production assessment.