Virtual power plant (VPP) is the owner and/or manager of a group of distributed generations as well as some end consumers, which may have random behavior. In addition to the uncertainties associated with generation and demand, the market price is also a source of uncertainty in VPP scheduling. In this paper, a novel stochastic operating strategy is proposed for VPPs to participate in day‐ahead market, considering different uncertainties in generation, consumption, and electricity market price. Highly stochastic consumption of plug‐in hybrid electric vehicles (PHEVs) is also considered in the present study, as an emerging source of uncertainty. To mitigate the loading effect of PHEVs on the network, a smart charging strategy is proposed, which encourages the owners to charge their PHEVs in off‐peak hours, and the results of this strategy are used in estimating the load. To achieve the maximum profit for the VPP in an electricity market, different methods based on point estimation method and Monte Carlo simulation are analyzed to handle the uncertainties. Moreover, three strategies are proposed for charging PHEVs, and their effect on VPP's total cost is compared. The proposed stochastic programming is solved by a modified version of teaching–learning‐based optimization algorithm. Numerical simulations on modified 18‐bus distribution system corroborate the efficacy of the proposed methodology. Copyright © 2015 John Wiley & Sons, Ltd.