Two experiments combined masked priming with ERP recordings in order to examine the effects of primes that are orthographic neighbors of target words. Experiment 1 compared effects of repetition primes with effects of primes that were high-frequency orthographic neighbors of low-frequency targets (e.g., four - tour). The prime was presented during 78 ms in a semantic categorization task, and Experiment 2 compared the same word neighbor primes with nonword neighbor primes (e.g., lude - luge). The prime was presented during 48 ms in a lexical decision task. The first observed effect occurred around 250 ms post-target onset suggesting an orthographic sub-lexical processing. At this stage, only the degree of orthographic overlap seems to influence word recognition processing. The second observed effect took place around 300 ms highlighting the lateral inhibition mechanism at the lexical level. This effect continued to the N400 component, reflecting the interactions between levels of representation for whole-words and concepts. The results are discussed with respect to possible mechanisms of lexical competition during visual word recognition.