Background: We aimed to propose a correction of the Lawton instrumental activity of daily living (IADL) score to take into account the possibility to have never done some activities, and measured its agreement and reliability with the usual IADL score. Methods: A cross-sectional study was conducted in outpatients attending French memory clinics between 2014 and 2017. Lawton IADL, cognitive performance, diagnosis, neuropsychiatric symptoms, and sociodemographics characteristics were collected. A corrected IADL was calculated and its agreement with the usual IADL was assessed. Results: The study included of 2391 patients (79.9 years old, 61.7% female). Based on the usual IADL, 36.9% of patients had never carried out at least one of the activities. This proportion reached 68.8% for men and 17.7% for women. Women had a mean IADL higher than men: 4.72 vs 3.49, this difference decreased when considering the corrected IADL: 4.82 vs 4.26 respectively. Based on Bland-Altman method, 93.5% of observations lied within the limits agreement. The ICC between the 2 scores was 0.98. The relationships between patients' characteristics and the IADL scores were similar, regardless the usual or corrected version. Conclusions: This corrected IADL score had an excellent degree of agreement with the usual version based the ICC. This simple correction could benefit both for the clinical practice by providing a more accurate description of the real clinical state of the patients allowing to manage them more precisely, and for research involving the evaluation of the functional abilities of patients.