BACKGROUND: Aim of this study was to detect predictors of better adherence to the AgeWell.de-intervention, a two-year randomized multi-domain lifestyle intervention against cognitive decline. METHODS: Data of 317 intervention group-participants comprising a risk group for dementia (Cardiovascular Risk Factors, Ageing and Dementia (CAIDE) score of  ≥ 9; mean age 68.9 years, 49.5% women) from the AgeWell.de intervention study were analysed. Regression models with four blocks of predictors (sociodemographic, cognitive and psychosocial, lifestyle factors and chronic conditions) were run on adherence to the components of nutrition, enhancement of social and physical activity and cognitive training. Adherence to each component was operationalised by assessing the degree of goal achievement per component at up to seven time points during the intervention period, measured using a 5-point Likert scale (mean score of goal achievement). RESULTS: Increasing age was negatively associated with adherence, while higher education positively predicted adherence. Participants with better mental state (Montreal Cognitive Assessment (MoCA)-score > 25) at baseline and higher self-efficacy adhered better. Diabetes and cardiovascular conditions were not associated with adherence, whereas smoking negatively affected adherence. Highest education and quitting smoking in the past were the only predictors associated with all four intervention components. CONCLUSION: Results identified predictors for better and worse adherence. Particularly self-efficacy seems to be of considerable influence on adherence. This should be considered when designing future intervention trials. TRIAL REGISTRATION: German Clinical Trials Register (ref. number: DRKS00013555).