Abstarct: Today, the optimal use of energy resources has become essential due to the increasing demand, limited fossil fuels, and environmental pollution. One way to maximize deploying existing resources is to integrate them into a unit called energy hub. Energy hubs allow the conversion and collaboration of different forms of energy, and result in increasing of efficiency, reliability, security, and network flexibility. In this dissertation, the concept of energy hubs is used to probabilistic optimal energy flow across energy networks. here, the objective functions are minimizing the energy costs and amount of emission. Inside the energy hubs, CHP, furnace, heat exchanger and transformer is used to transfer and convert various energies. Here, networks include electricity, natural gas and heating. The considered uncertainty is related to network demands, which is implemented by the two-point estimation method. First, the impact of considering energy hubs on probabilistic energy flow in the networks is investigated. Considering energy hubs can reduce the energy costs, emissions and risk of operation. In order to validate the results, the utilized uncertainty method is compared with Monte Carlo simulation. Simulations have been carried out on the modified IEEE 14-bus electrical network, 11-node natural gas network, 11-node heating network and 11 energy hubs as well. The optimization problem here is non-linear and has been implemented using MATLAB and GAMS software.