BACKGROUND: Computerized cognitive training conducted in a home setting has shown beneficial effects on cognitive functions in people with mild cognitive impairment. Similar effects in people suffering from Alzheimer's disease (AD) have not yet been found. We aimed to examine the effect of a six-month tablet-based multimodal training in community-dwelling people with mild to moderate AD on cognitive functions and on the volume of (sub)cortical structures. METHODS: Within a randomized controlled trial, a six-month (un)supervised tablet-based multimodal training including cognitive and physical exercises in people with mild to moderate AD (n = 11) was compared to a control group (n = 11) that received cognitive paper-and-pencil exercises for voluntary, unsupervised training. Participants in the intervention group were visited by professional trainers on a weekly basis for joint, supervised training sessions and were encouraged to train alone or with a caregiver as often as possible. Neuropsychological examination included assessments of global cognitive functions, memory, attention, executive functions, and verbal fluency. Freesurfer analyses of T1-weighted scans from structural magnetic resonance imaging were used to assess volumes of specific (sub)cortical areas (e.g., hippocampal volume). RESULTS: Over six months, the intervention group showed a stable global cognitive function score (Mini Mental Status Examination), whereas the control group showed a cognitive decline (ANCOVA-interaction: F1, 14 = 5.083, p = .041; controlled for disease duration and education). No other selective cognitive domain showed a significant time-by-group difference. No difference in cerebral volumes were detected. CONCLUSION: The tablet-based multimodal training with cognitive and physical activation has positive effects on global cognitive functions of people with mild to moderate AD over a six-month training period, but lacks measurable transferability to other cognitive domains such as memory, attention or executive functions or brain structure. Further research on such interventions using high-quality longitudinal designs is recommended. TRIAL REGISTRATION: ClinicalTrials.gov (NCT04628702).