Previous studies have found viewpoint-dependent effect in object recognition. So one important question is whether view-dependent effect is held in object search. The present study explores the view-dependent effect and its roles in real object search. The pilot study was used to select 25 objects from 60 objects. Each type of object has two pictures-one canonical view and one accidental view and totally 50 pictures was used as formal experimental material. Real object search paradigm and two (cue type: picture cue vs. name cue) by two (target object view: canonical view vs. accidental view) within-subject design were employed. The events of one trials were as follows: The fixation across of 500-800 ms was firstly presented at the center of screen. Then a picture or a name cue of target object was remained on screen for 1000 ms and was followed by an 800-ms blank screen. Then search display composed of six items (one target object and five distractive objects) was presented and participants were asked to search target object. When participants find the target object, fixate it and press the response key. The picture cue and name cue were counterbalanced in blocks between participants. The whole experiment composed of two blocks of 50 trials (totally 100 trials). The experimental stimuli were presented using a PsychoPy-programmed PC and LCD monitor (21" TFT; resolution: 1920 ×1280 pixels; refresh rate: 60 Hz; view distance: 70 cm). The SMI RED eye tracking system (SensoMotoric Instruments, Teltow, Germany) were employed to record the participants’ eye movements during the whole search processes. As shown in Figure 1, the findings of this study are as follows: A similar data pattern appears on accuracy and response times, showing a significant interaction between cue type and view type of target object. Specifically speaking, when the search cue is the name of target object, the search performance for canonical view target object is superior to that for accidental view of target object (accuracy: p = 0.014; reaction times: p < 0.001); when the search cue is the picture of target object, there is no difference in search performance between canonical view and accidental view conditions (ps > 0.210). A similar interactive data pattern also appears on eye movements during scanning epoch and verification epoch of the whole search processes. Specifically speaking, when the search cue is the name of target object, the durations of scanning epoch and verification epoch for canonical view target object are short than those for accidental view (ps < 0.001), and the fixation numbers of scanning epoch and verification epoch for canonical view target object are less than those for accidental view (ps < 0.001); when the search cue is the picture of target object, there is no difference in durations and fixation numbers of scanning epoch and verification epoch between canonical view and accidental view conditions (ps > 0.05). It can draw the following conclusions: in the search for object with the name as cue, the canonical view enhances the efficiency of attentional guidance by select the potential locations of the target object and speeds up the target object verification process by reducing the fixation number. So the view dependent effects are held in both scanning epoch and verification epoch, supporting for the dual-function theory of target template