Previous studies on the effect of reward on selective attention mainly focused on reward attributes, reward learning and reward history, but less attention was paid to the effect of anticipated stage of reward on selective attention.Reward anticipation refers to the desire for the upcoming reward. The neuron activation involved in reward anticipation stage continues until the reward is obtained. In the reward anticipation stage, the reward can induce the individual to have a strong motivation to complete the task.This study is to explore the effect of reward anticipation on selective attention. We carried out two experiments combining monetary incentive delay task and two-rectangle paradigm.A cue indicating the reward condition of each trial (incentive vs. non-incentive) or (low-incentive vs. high-incentive) was followed by the presentation of two-rectangle paradigm. participants were asked to make a key response to the goal in the two-rectangle paradigm quickly and accurately. the goals included four kinds of situations. "valid": the target was in the cued location (96 trials), "invalid same-object": the target was at the uncued end of the cued rectangle (32 trials), "invalid different-object":the target was on the uncued rectangle (32 trials) and "capture stimulus":the target was absent (64 trials). If the target didn’t appear, there’s no need to make any response. The reward and non-reward conditions were equally divided, and the trials were presented randomly. After the completion of the experiment, the participants were given a basic reward of 10 yuan and additional reward based on performance. Reaction times and the error rates were recorded. After repeated measures of variance analysis of the obtained data, we found that in experiment 1, the main effect of reward anticipation is significant ( F (1,29) =4.25, p =.048, η2 =.13), under the condition of anticipated reward, reaction time is significantly faster than no anticipated reward conditions. The interaction of reward anticipation and the validity is marginal significant (F(2,28)=2.71，p =.08, η2=.09). Under the condition of reward anticipation, the response time under the condition of valid is significantly lower than that under the condition of invalid same-object and invalid different-object (space-based attention exists), and the RT under the condition of invalid same-object is equal to that under the condition of invalid different-object (object-based attention almost doesn’t exist). Under the condition of no reward anticipation, the response time under the condition of valid is significantly lower than that under the conditions of invalid same-object and invalid different-object , and the response time under the condition of invalid same-object is significantly lower than that under the condition of invalid different-object. In experiment 2, the main effect of reward anticipation is not significant (F (1,22) =.33, p =.57). The interaction of reward anticipation and the validity is significant(F (2,21) = 1.57, p=.02, η2=.18).Under the condition of high reward anticipation,the response time under the condition of valid is significantly lower than that under the condition of invalid same-object and invalid different-object, and the RT under the condition of invalid same-object is equal to that under the condition of invalid different-object . Under the condition of low reward anticipation, the response time under the condition of valid is significantly lower than that under the conditions of invalid same-object and invalid different-object , and the response time under the condition of invalid same-object is significantly lower than that under the condition of invalid different-object . These results show that reward anticipation affects selective attention and has different effects on object-based and space-based selective attention.