The present study investigates the effects of neighborhood size and neighborhood frequency in reading Chinese two-character words. The neighborhood size of a word is defined as the summation of neighbors sharing the first constituent (neighborhood size 1) and the second constituent (neighborhood size 2) characters. The first experiment found two opposite neighborhood size effects in lexical decision of high-frequency and low-frequency words. The regression analysis showed that neighborhood size 1 influenced word reading more than the neighborhood size 2. The second experiment confirmed this finding and showed that reading words with higher frequency neighbors took a longer time and elicited greater N400 and LPC than those without higher frequency neighbors. These findings indicate that, when reading Chinese two-character words, all words sharing the first constituent character are activated in the early stage of word recognition and the existence of high-frequency words among neighbors leads to greater competition in the stage of semantic integration and response selection.