Cognitive aging has become a major concern because life expectancy has increased and elderly populations are socially and economically active. Neurofeedback is a technique of neuromodulation through operant conditioning paradigm that uses a computer interface to provide real-time information about brain activity to increase individual self-perception and assist in modulation. The sensorimotor rhythm (SMR) training protocol is known to enhance attention and has been applied to improve cognitive performance, primarily for attention and memory gains. The aim of this study is to test if the SMR protocol can improve working memory performance in an aging population and consequently favor cognitive reserve. Seventeen older adults (12 females) took part in a randomized placebo-controlled study. They completed a visual working memory test, Delayed Matching to Sample Task (DMTS), before and after the SMR neurofeedback protocol in order to compare their visual working memory performance. Moreover, a 19-channels EEG was collected while they perform the DMTS pre- and post-training. The experimental group showed an improvement in their working memory performance after the training with similar activation power, mainly in theta and beta frequency band at frontal and alpha at temporal regions. The sham group showed some variations in the score of working memory after the training, but were not statistically significant and their power spectrum demonstrate enhancement in alpha and beta band frontal and temporal. The group that did not receive neurofeedback training did not show a change in their working memory performance, neither in their EEG spectrum. The results suggest that neurofeedback can benefit brain reserve in an aging population because individuals enhanced their working memory performance after training and have their EEG activation changed according to expected in working memory tasks.