This study investigates how electrode number and spatial configuration affect electroencephalography (EEG) source localization performance in the context of visual evoked potentials (VEPs). Five healthy participants performed visual stimulation tasks while EEG signals were recorded with a 256-channel cap. Six electrode montages were evaluated: four whole-head configurations (32, 64, 128, and 256 channels) and two targeted layouts (an 84-channel occipital-only montage and a 164-channel montage excluding frontal sensors). Source localization performance was assessed based on VEP neurophysiological expectations using three complementary indices: a Lateralization Index (LI) for hemispheric dominance, a Signal-to-Noise Ratio (SNR) at the source-space level, and a custom Occipital Precision Index (OPI) that quantifies how tightly activation is confined to occipital regions. Higher-density montages (128 and 256 channels) achieved the highest OPI values (≥0.68), yielding the most focal localization and reflecting accurate activity confinement to the visual cortex. Notably, the targeted 84-channel occipital montage performed comparably to both the 128-channel whole-head and the 164-channel targeted configurations in OPI and SNR, demonstrating that dense sampling over the region of interest can rival broader higher-density coverage. In contrast, lower-density whole-head montages (≤64 channels) exhibited inflated LI and SNR values but lower OPI, indicating reduced spatial precision despite seemingly higher signal metrics. The findings suggest that targeted electrode configurations can approach the performance of high-density caps when the region of interest is known in advance, though the small sample size (N=5) warrants caution and further validation.