Background and Objective: Dementia is a chronic progressive disease characterized by multifactorial aetiology. Among the factors of greatest interest for prevention is lifestyle, including dietary habits. In particular, tea and coffee are still a matter of debate, being associated both with protective and detrimental effects. The aim of our study is to assess the relation between tea and coffee intake and risk of developing dementia through a dose-response meta-analysis based on cohort studies. Methods: We performed a literature search in PubMed and EMBASE databases up to 28 November 2022. Inclusion criteria were: population without chronic diseases or previous dementia diagnosis, cohort study design, assessment of tea or coffee intake by increasing categories of exposure, assessment of incident dementia risk. Exclusion criteria were: type of study (non-cohort), type of publication (letters, commentaries and systematic reviews), missing information about exposure level for tea and coffee in each category. For statistical analysis, we used cubic splines through a non-linear random-effects regression model (drmeta routine in Stata/SE 17.0). Results: We identified 9 articles published between 2009-2021, 1 assessing tea consumption, 6 coffee and 2 both. As the amount of tea consumed increases, risk of dementia decreases progressively. Conversely, we found a U-shaped association between coffee intake and dementia risk: as the number of cups of coffee consumed increases, the risk of dementia decreases progressively up to two cups of coffee (approx. 300 mL) per day, whereas as the daily amount increases, there is a reversal up to five cups (approx. 750 mL) per day. Conclusions: Moderate coffee intake appears to have a protective effect, while intake in higher amounts seems to increase the risk of dementia. Concerning tea, the relation is almost linear in a protective way.