GUM is a common method for uncertainty evaluation. Inclusion intervals are calculated by propagating uncertainty with GUM, which can get accurate evaluation results. Uncertainties of meteorological instruments verification results are also evaluated by GUM. However, GUM results may appear deviation when measurement model and calculations are complex. Monte Carlo Method (MCM) provides a general numerical method for uncertainty evaluations by propagating probability distribution. Take verification results of air pressure sensors as examples, uncertainty was evaluated by MCM and GUM results were verificated. Results showed that GUM results did not pass the verification. There are some problems in evaluating air pressure uncertainty with GUM. A more accurate inclusion interval can be calculated by MCM.