BACKGROUND AND OBJECTIVES: Cognitive behavioral therapy (CBT) is a first-line treatment for anxiety, but it is not widely available as clinical guidelines recommend. We examined the feasibility and efficacy of a novel smartphone-based fully automated digital CBT intervention, 'Daylight™', to improve symptoms of Generalized Anxiety Disorder (GAD). METHODS: In this multiple-baseline design, 21 adults (20 F; mean age 43yrs. range 19-65yrs.) with moderate-to-severe symptoms of GAD were randomized to one of three baseline durations (2-, 4-, or 6-weeks) and then received access to digital CBT. Participants completed daily ratings of anxiety and worry, weekly measures of anxiety, depressive symptoms, and sleep, and measures of anxiety, worry, wellbeing, quality of life, CBT skill acquisition, and work performance at initial assessment prior to baseline randomization, post-intervention, and follow-up. RESULTS: Digital CBT was found to be feasible in terms of engagement, satisfaction, and safety. For preliminary efficacy, improvements were detected in daily and weekly outcomes of anxiety for most participants. Despite individual differences, significant improvements occurred with the introduction of digital CBT and not during baseline. Overall, 70% of participants no longer had clinically significant symptoms of GAD, 61% no longer had significant depressive symptoms, and 40% no longer had significant sleep difficulty at post-intervention. LIMITATIONS: The study sample was recruited using the internet and was mostly female, limiting the generalizability of the findings. CONCLUSIONS: Findings support the feasibility and efficacy of Daylight. Further examination in randomized controlled trials is now warranted.