The aim of this study is to make the recall-based reading span test(Daneman & Carpenter, 1980) into the recognition-based test for on-line measuring the capacity of the participant. In order to measure the concurrent validity, a series of experiments is performed with varying features of distractors consisting of the reading span test. In experiment 1, which included irrelevant words as distractor, low correlation was observed. And in experiment 2, including several types of distractors which interfere with the selection of target words, low correlation was observed too. But in experiment 3, including distractors no more relevant in the Present trial but relevant in previous trial, high correlation was observed. The results of this study have theoretical implications on the validity of the reading span test and practical implication in that this study provides the tool for the studies on individual differences in working memory capacity.