To investigate the time course of instant foveal priming during reading, Sereno & Rayner (1992) designed the fast priming paradigm. The present paper introduced the fast priming paradigm in terms of basic principles, design backgrounds, application fields and matters needing attention. Furthermore, the application prospects of fast priming paradigm in Chinese reading are also discussed. As a kind of eye movement research paradigm, fast priming paradigm is also based on eye-movement contingent display-changes technology. In this paradigm, a string of random letters initially occupies a specified target word location as a participant reads a sentence. When the reader's eye crosses an invisible boundary located at the left boundary of the target area, the random-letter string is replaced by a prime word for a specified and brief duration. During the fixation on the target region, the prime word is replaced by the target word, which remains in place while the subject finishes reading the sentence. Thus the fixation time on the target word was considered as a function of the relation between the prime and target. The duration of the prime was varied to determine the optimal level of priming. Indeed, fast-priming paradigm is similar in general to masked priming paradigm (Forster & Davis, 1984). The major advantage of this paradigm is that it can be used to study word identification in the silent reading of text with minimal disruption of normal reading. A second advantage is that the fixation times obtained are appreciably shorter than manual or vocal latencies. By precisely manipulating prime types and prime durations, fast priming paradigm can examine the time course of priming for lexical information (e.g., word frequency, word length) and sublexical information (e.g., phonological, semantic and orthographic codes) during word recognition. Thus it provides a valuable new tool for verifying and improving related theories of lexical structure and lexical processing and a feasible new approach to investigate special research questions such as lexical ambiguity. However, it is necessary to firstly identify the factors that constrain the presentation of fast priming. Firstly, the prime duration must be brief. Secondly, both the predictability of target words and spillover effect should be controlled. Thirdly, the prime and also the target words should be relatively short to minimize the number of multiple fixations on the target. Fourthly, as there are two changes of stimuli on the target word location, the refresh rate of display screen must turned high. Besides, the standands of data processing are also worthwhile to note. It has been more than 20 years since the fast priming paradigm was designed. The studies using this paradigm are limited, especially in Chinese reading, though the application prospect is really broad. For example, to extend the concept of sublexical information to Chinese, the roles of semantic radicals and phonetic radicals in recognizing the character can be examined. Besides, we can also investigate the distinction between onset and rime in Chinese syllable. In summary, a better understanding of the fast priming paradigm must provide researchers an effective approach and some new ideas for research questions.