In light of demographic change, age-related cognitive decline poses a significant economic and public health challenge. Non-pharmacological interventions, such as engaging with information and communication technologies, offer promising avenues to mitigate cognitive decline. Existing evidence links internet use among older adults to positive outcomes such as improved mental health, social connectedness, and cognitive function. However, prior research has often overlooked the potential for bias introduced by learning effects in repeated cognitive testing, potentially inflating reported associations. This study empirically examines the link between internet use and cognitive functioning in individuals aged 65 and older, using data from four waves of the Survey of Health, Ageing, and Retirement in Europe. Employing linear mixed models and multilevel logistic regressions, we account for potential learning effects and resampling bias. Our findings partly underscore the established association between internet use and certain cognitive measures, offering insights into bias in repeated cognitive sampling.