The invention discloses a fingerprint database construction method of a WiFi indoor positioning system. The method comprises the following steps: determining the position and the coordinates of a reference point, capturing the media access control address of access points at the reference point, collecting signal strength of all the access points in a media access control address list, and sending a sampling file to a server; conducting statistical treatment on signal strength data in the sampling file to obtain a fitting probability distribution curve, conducting correlation correction on the fitting probability distribution curve, conducting truncation processing on the probability distribution curve obtained after the correlation correction, calculating the probability of each decibel milliwatt value, conducting normalization processing, and finally generating a fingerprint database. The method solves the problems that the establishment process of a traditional fingerprint database is longer in consumed time, the processing is tedious, the fingerprint information is insufficient, and updating and maintenance are not facilitated. Due to the parallelization of fingerprint collection, the establishment of the fingerprint database is accelerated, optimized processing is carried out on the collected fingerprint data, and the fingerprint database is more accurate and comprehensive.