Introduction: Bilingual researches about languages switching focused on three fundamental issues in the processing of the non-target language in language switching: (1) whether the non-target language was activated when bilinguals intended to use the target language? (2) If the non-target language was activated, then at what level did this activation occur, at the conceptual level or the lexical level? (3) If the non-target language was activated, then what influence would it exert on the processing of the target language? As the first two questions have been answered, we try to answer the third question.Used cross-language instant semantic priming and long-term priming, under semantic decision task (Biological properties judgment) to investigate how non-target language affect target language at the time courses for the late fluent bilinguals. Design: Chinese as a target language and English as a non-target language in experiment 1;English as a target language and Chinese as a non-target language in experiment 2. Method: Twenty-three native Chinese/English bilinguals (21-25 years old) participated in experiment1 and twenty-four participants (21-25 years old) took part in experiment2. Semantic relatedness was controlled based on a rating procedure conducted with 20 native Chinese/English bilinguals using a Lickert scale (1 = unrelated, 5 = strongly related). Semantically related pairs were significantly more related than semantically unrelated pairs (all ps < 0.001). All words were presented one after the other at fixation for 1000 ms. Block order and response sides were counter balanced to cancel out carry over effects. Participants were asked to finish Biological properties judgment in the experiment, which contents two parts: instant semantic priming and long-term priming. Result: In experiment 1, Semantically related pairs and semantically unrelated pairs were not significantly different neither in instant semantic priming(t =-0.423，p =0.676) nor long-term priming(t =-0.82，p =0.419).In experiment 2, semantic short-term priming from L1 to L2 was significant(t =-3.05，p =0.006) while semantic long-term priming was not (t =-0.63, p =0.536). Conclusion: The two experiments’results show that instant priming occurred only when English as target language and Chinese as non-target language; cross-language long-term priming did not occur in experiment 1 nor experiment 2. There is an asymmetry of semantic priming in late fluent bilinguals.