This article takes SOFC-PEM hybrid systems as the research object and considers the operating parameters to be uncertainty. Uncertainty analysis method based on Latin hypercube sampling involves quantifying the uncertainty in the input parameters in the form of appropriate distribution functions, propagating the uncertainty through a deterministic model to construct the output variability distributions, and analyzing the effects of operational parameter uncertainty on system output. Furthermore, non-traditional sensitivity analysis approach is used to evaluate the relative importance of various parameters considering this distribution structure. Results show that higher operating temperature of SOFC stack, lower fuel rate and lower fuel utilization of SOFC stack can reduce the influence of uncertainty on system output at the expense of system output. Uncertainty analysis can provides more credible basis for the optimization of design parameters under uncertainty.