Electroencephalographic neurofeedback (EEG-NF) protocols commonly present visual stimuli, modified at fast frequency-rates, in order to inform participants about their EEG signal fluctuations in real-time. The current study addresses methodological issues in understanding the resulting neuromodulation induced by these protocols. Specifically, we investigate whether classically targeted EEG features: 1) tend to fluctuate naturally over time, 2) are influenced by the presentation of a continuously modified visual stimulus and by 3) the frequency at which the stimulus is modified. We enroll 32 healthy young adults in a classic EEG-NF environment without submitting them to a NF task (i.e., modulation of brain activity). Participants only have to stare at a gray circle presented at the center of a screen during EEG recording at Fz, Cz, and Pz scalp positions (10-20 international system). Participants complete four blocks of eight one-minute trials each. In one block, the circle remains the same across trials. In the remaining three blocks, the circle size randomly changes at different frequencies depending on the block: 1, 5, or 10 Hz. Results indicate a natural increase in alpha power across trials, regardless of scalp position, and an increase in theta power at Pz when the circle size changes continuously. These findings have implications for theta and alpha EEG-NF, suggesting that the observed neuromodulation during these protocols may in part arise from non-specific factors inherent in the protocols design. This understanding calls for caution when interpreting the effects of EEG-NF protocols on EEG signal. Methodological considerations prove essential for future research.