A complete uncertainty analysis on a given fringe projection set-up has been performed using Monte-Carlo approach. In particular the calibration procedure is taken into account. Two applications are given: at a macroscopic scale, phase noise is predominant whilst at microscopic scale, both phase noise and calibration errors are important. Finally, uncertainty found at macroscopic scale is close to some experimental tests (~100 {\mu}m).