ABSTRACT Introduction Ambulatory assessment of executive function – particularly in the form working memory (WM) – is increasingly common. Few studies to date, however, have also incorporated ambulatory measures of inhibitory control. Critically, the extended within-person reliability of ambulatory tasks tapping each of these constructs has been largely overlooked. Method Participants (N = 283, M age = 23.74 years, SD = 9.04) received notifications every 3 days (for 4 weeks) to undertake ambulatory assessment versions of the n-Back and Stop-Signal Tasks (SST) via the smartphone application CheckCog. Within-person reliability of these measures was explored. Results Compliance ranged from 66% (for eight sessions) to 89% (for four sessions). Our results reveal significant changes in performance within the first two sessions for both the n-Back and SST, with performance remaining largely consistent across the remaining (two to eight) sessions. In terms of test–retest reliability, the ICC (C, 1) values ranged from .29 to .68 on the n-Back (with overall accuracy being .51) and .31–.73 on the SST (with stop-signal reaction time being .53). Conclusion The results of the current study contribute to the literature by demonstrating the reliability of brief measures of executive function – in the form of inhibitory control and WM – delivered using smartphones in participants’ natural environments. Based on our findings, the CheckCog app reliability tracks baseline systematic changes in WM and response inhibition across multiple time points and for an extended period in healthy individuals.