Abstract There are two different explanations for the Attentional Blink(AB)—a deficit in reporting the second of two targets when presented in close temporal succession: one is central capacity limitations, which assumes that the AB reflects an unavoidable resources limitation; the other is resources allocation, which describes the AB as a product of overinvestment of attentional resources. If the AB is determined by an allocation policy, it could be mediated not only through top-down task performances, but also by bottom-up stimulus variations. Although the flurry of recent findings showing that the AB can be attenuated in various mission requirements, few papers show that task-irrelevant stimulus character can attenuate the AB as well. In the present study, we tested the hypothesis of novel stimulus-driven Attentional Capture (AC) regarding the resources allocation mechanisms by which stimulus circumstances result in a reduced AB effect. An AB task varying in 4 color types was used. Type 1 was a standard AB task in which both the first target (T1) and the second (T2) were digits in an RSVP stream consisting of 18 interference letters, all black on a white background. Type 2-4 were largely the same as type 1 except for the following changes: both the T2 and T2-1(interference letter previous to T2) were red in type 2; the T2 in type 3 and T2-1 in type 4 were red too; the T2+1(interference stimulus next to T2, also named mask stimulus) kept the same color with T2 in 4 types respectively. In each type, the temporal distance between T1 and T2 was systematically varied in 2, 3, and 8 items (Lags 2, 3, and 8). Stimulus generation and response recording were done with E-Prime2.0. Each letter or digit was presented for 30ms, followed by a 70ms blank. Forty-six participants completed all 4 types. Their task was to identify both T1 and T2. An unspeeded response was made at the end of each trial by typing the digits on a standard keyboard. During the test, they were asked to ignore the interference letters or any possible changes in colors. Accuracy data were submitted to an analysis of repeated measuring variance with color type (1, 2, 3, and 4) and lag (2, 3, and 8) as variables. The results for T1 showed that color type and lag had no significant effects on AB task. The results for T2 when T1 was identified correctly showed that there was no interaction effect between color type and lag, F (6, 270) = .895, p = .503. The main effect of lag was significant, F(2,90) = 9.665，p < .001. Accuracy dropped after Lag 2 and then gradually increased again. There was a main effect of color type, F(3,135) = 4.405，p < .01. Performance was better overall in types 2, 3, and 4 than in type 1, F(1,45) = 11.576，p < .01, but there were no significant differences among types 2, 3, and 4, F(1,45) = 2.378，p = .130. These findings support the hypothesis that the AB is due to an allocation of limited attentional resources in stimulus processing. The patterns of resources allocation can be mediated by AC automatically. Driven by task-irrelevant novel information on T2-1 or T2, the AC can leave the excess of resources that would normally be allocated to T1 available for T2.