Important engineered slopes are often heavily instrumented and their performance routinely monitored through these instruments. The evaluation of the safety of the slopes based on the monitored information is however a challenge. A systematic method is presented in this paper for evaluating the slope safety by combining multi-source monitoring information with underlying physical mechanisms. First, a Bayesian network with continuously distributed variables for a slope involving the factor of safety, multiple monitoring indexes and their influencing soil or rock model parameters is constructed. Then the prior probabilities for the Bayesian network are quantified considering model and parameter uncertainties. After that, multi-source monitoring information is used to update the probability distributions of the soil or rock model parameters and the factor of safety or failure probability using Markov Chain Monte Carlo simulation. Two rock slope examples are worked out to illustrate the proposed methodology. A non-intrusive stochastic numerical method is used in the reliability analysis in the examples.