Researchers believe that learning a second language can promote the bilingual cognitive control skills, which may also reduce the ability of bilingual lexical access. But there is a contradiction in many previous literatures and whether or not there exists an bilingual cognitive control advantage is no sure. In this study, we choose the SOA paradigm Stroop task to investigate the timing and magnitude of Stroop interference and facilitation in bilinguals to compare the differences between cognitive control abilities and lexical access. Our hypothesis is that if bilinguals had not cognitive control advantages, on larger values of SOA the reaction time of interference effect would longer in second language, that is , there was a lexical access disadvantage, bilinguals’ interference effect in first language would have a longer reaction time comparing to the second language. Sixty-one college students participated in the experiment, whose average age was 21.62 years old. Experiments used 3 (whether color word meaning is consistent with the color block’s color or not: consistent, inconsistent, control) ×5 (SOAs: -400ms, -300ms, -200ms, -100ms, 0ms) two-factor within the design. Participants completed Chinese and English language experiment. We used “红 绿 蓝”and “red green blue” as word stimuli. Colored rectangles were red, green or blue. In negative SOAs, the word stimulus appeared on the screen alone first for either 400ms, 300ms, 200ms, or 100ms, followed by the colored rectangle. In the 0ms SOA, the word stimulus and the colored rectangle appeared simultaneously. Both stimuli remained on the screen until participants made a response, if no response was made in 2000ms, the stimuli disappeared and the next trial began. We used E-prime software programming and SPSS16.0 analyzing data. The results show that in L1 Chinese participants’ peak interference effect in the Stroop task occurred at -200ms SOA, which is inconsistent with the result of -200ms SOA(Coderre,et.al.,2013) and it may because of the different language environment. In L2 English conditions, the experiment measured largest Stroop interference effect and it was caused by the -300ms SOA, which means that the time appeared negative transform in interference effect. It’s inconsistent with 0ms SOA (Glaser et.al., 1982) but has the same change trend with -200ms SOA(Coderre,et.al.,2013),which may because of different response ways. In L2 English conditions, -400ms SOA triggered the largest facilitation effect while under the L1 Chinese conditions the largest facilitation effect occurred in -300ms SOA, which prove there is a bilingual lexical access delay. Moreover, there is not significant difference in interference effect between Chinese and English language environment, and marginal significant in facilitation effect between the two language. The data shows that Chinese and English has different language script that may influence the performance of facilitation effect. And Chinese words have spatial properties so that participants may spend more time to recognize the words comparing to English words. In this study, the bilingual cognitive advantage was not observed in the experiment, and bilinguals were affected by lexical access abilities, which support the bilingual lexical disadvantage hypothesis. And second language learning may reduce the lexical access of the bilinguals’ mother tongue.