OBJECTIVE: Verbal fluency tasks are principally used to assess lexical access and have shown usefulness for differential diagnosis. The purpose of Study 1 was to provide normative data in the adult French-Quebec population (Canada) for semantic verbal fluency (animals), for two sets of phonemic verbal fluency (TNP and PFL), and for letter P alone (60 seconds per category/letter). The objectives of Study 2 were to establish the diagnostic and predictive validity of the present tasks and normative data in Alzheimer’s disease (AD) and major depressive episode (MDE). METHOD: The normative sample consisted of 932 participants aged 19-91 years. Based on multiple linear regressions, equations to calculate Z-scores were provided. To assess validity, performance of 62 healthy participants was compared to 62 participants with AD and 41 with MDE aged over 50. RESULTS: Age and education, but not gender, predicted performance on each verbal fluency task. Healthy adults aged 50 and younger had a better performance on semantic than phonemic verbal fluency. In comparison to MDE, AD participants had lower performance on animals and TNP, but not on letter P. Ninety percent of people with a Z-score </= -1.50 on semantic verbal fluency had AD and the global accuracy was 76.6%. Test-retest reliability over one year was high for both animals (r = .711) and TNP (r = .790) in healthy older participants, but dropped for animals in people with AD (r = .493). CONCLUSIONS: These data will strengthen accurate detection of verbal fluency deficits in French-Quebec adults.