Much evidence suggests that low-level visual variables rather than high-level linguistic variables affect initial landing positions in alphabetic language scripts reading (Rayner, 2009). Written Chinese differs from alphabetic writing systems in many dimensions. Chinese uses a nonalphabetic, character-based script with square-shaped forms of different levels of visual complexity (i.e., roughly the number of strokes) as basic writing units. Yang and McConkie (1999) obtained a flat preferred viewing location curve and concluded that, unlike other languages, saccadic target selection is not word-based in Chinese reading. The results also showed that the number of strokes did not influence participants’ initial landing positions. However, we think that, in their study, the width of the space between adjacent characters is equivalent to half the width of a Chinese character, which maybe influence the results. Many researches have found that the number of strokes influences the processing of vocabulary recognition. The results suggested that readers tended to fixate the more complexity characters. In current study, sentences including two-character target words were adopted to examine whether the number of stroke influenced the landing positions. The number of stroke of the first and the second constituent characters were manipulated. 60 participants’ eye movements were monitored as they read texts. Their eye movements were recorded by a SR Research EyeLink 2000 eyetracker (sampling rate = 1000 Hz) that monitored the position of the right eye every two milliseconds. We found that there was different eye movements behavior in different fixation cases. When there was only one fixation on target word, the first fixation mostly landed on the centre of the words. While there were multiple fixations, readers first fixated at the beginning of the target words. There was a preferred viewing location in single-fixation cases during Chinese reading. In multiple fixation cases, if the first fixation landed at the beginning of a target word, the probability of refixating this word was the highest. Importantly, we found that when the number of first character’s strokes is high, readers mostly fixated the first character. And both the number of strokes of the first and second characters influence the probability of refixation. We argued that Chinese children use the “strategy-tactics” approach during reading.