Neurofeedback training (NFT) has been proven to have positive effects for human cognitive performance enhancement and to be a potential non-pharmacological supportive treatment for many neurological and psychiatric disorders. The mental strategies the participants use during the NFT are a determinant feature in a successful NFT. However, most existing studies focus on either the training protocols or the training effects. Therefore, this study investigates the mental strategy effects during beta/theta ratio based NFT and attempts to find an efficient method to analyze the mental strategies that can be used to improve the success rate in NFT researches. Objective beta/theta ratio value scores and qualitative mental strategy classification were used for analysis. The results showed that no specific mental strategy is more significantly useful for NFT under this protocol but the strategies that required more attention performed better, which is consistent with the findings in previous NFT studies.