In sentence production, syntactic priming effect refers to a phenomenon that speakers or writers prefer to reuse syntactic structures that they just processed, and the magnitude is larger for less frequently used (non-preferred structure) than for more frequently used (preferred structure). Different theories have been proposed to account for the syntactic priming effect. The implicit learning theory assumes that the syntactic structure of prime sentences strengthens the connection between the specified syntactic structure and the specific content of a sentence. The two-stage theory hypothesizes that syntactic encoding involves syntactic structure's selection and content planning. The syntactic structure of a sentence is determined by the maximum activation of syntactic nodes, and the production latencies is determined by the activation difference of all activated syntactic nodes. Studies have addressed the behavioral responses including syntactic choices ratio and production latencies, however, little is known about the eye-tracking pattern in planning of sentence production. The present study investigated the eye-tracking pattern of syntactic priming effect in Chinese spoken sentence production. In a syntactic priming task, participated were presented a sentence (as a prime) auditorily, and then they need to judge a visually presented sentence matched the prime sentence or not. After that, participants were instructed to describe the target picture orally using the verb “give”. The eye-movement and production sentences were recorded. We manipulated prime sentence type (double objects, prepositional objects, and simple sentence) and target sentence type (double objects and prepositional objects). For the syntactic choices ration, results showed that the type of prime sentences affected the syntactic structure choice ratio, presenting that in comparison with simple sentence prime condition, speakers produced more double objects (or prepositional objects) sentences in the double objects (or prepositional objects) prime condition. For the eye-movement pattern, results showed that it was modulated by prime sentences, regardless of the consistency about syntactic structure between primes and targets. In the double objects prime condition, fixation ratio on indirect object was significantly higher than that of direct object, and fixation sequences contained more sequences of indirect object in priority, while in the prepositional objects prime condition, fixation ratio on direct object was significantly higher than that of indirect object, and fixation sequences contained more sequences of direct object in priority. By contrast, in simple sentence prime condition, the eye-movement pattern was determined by the type of target sentences. Fixation ratio on indirect object was significantly higher than that of direct object when producing targets with double objects, and fixation ratio on direct object was significantly higher than that of indirect object when producing targets with prepositional objects. Our findings indicated that there is a syntactic priming effect in sentence production, presented in syntactic choice ratio and eye-tracking pattern. We suggest that the findings on eye-tracking data provide evidence for the two-stage theory in sentence production.