Abstract
<jats:title>Abstract</jats:title> <jats:sec> <jats:title>Significance</jats:title> <jats:p>fNIRS is highly suitable for the study of reading development, however, the reliability of its signals is not well understood during reading tasks.</jats:p> </jats:sec> <jats:sec> <jats:title>Aim</jats:title> <jats:p>Therefore, this study assessed the test-retest reliability of the fNIRS signal during a common event-related reading paradigm.</jats:p> </jats:sec> <jats:sec> <jats:title>Approach</jats:title> <jats:p>English-speaking adults (n = 30) completed a lexical decision task during fNIRS recording twice, one week apart.</jats:p> </jats:sec> <jats:sec> <jats:title>Results</jats:title> <jats:p>Our results demonstrated contrast effects partially consistent with prior neuroimaging literature, insofar as for each contrast, at least one predicted region of interest was activated. However, we did not identify significant activation in all predicted brain areas. Regarding group level test-retest reliability, we observed poor reliability across predicted brain regions for most conditions, with the exception of fair test-retest reliability in the left posterior temporal lobe for coarse lexical tuning. At the single-subject level, test-retest reliability ranged from poor to excellent across subjects, but was poor for most subjects.</jats:p> </jats:sec> <jats:sec> <jats:title>Conclusion</jats:title> <jats:p>These results suggest fNIRS can detect changes in brain activation during a fast event-related reading task at the group level. However, reliability does not appear sufficient to interpret individual-level data. Further research should explore reliability across a wider range of designs to assess the generalizability of these findings.</jats:p> </jats:sec>