This paper presents a robust optimization-based decision-making tool for the daily and weekly self-scheduling of Virtual Power Plants (VPPs) in the uncertain environment of electricity markets. VPP, as a heterogeneous coalition of distributed energy resources (DERs), is generally composed of intermittent renewable sources, storage systems, flexible loads, and small conventional power plants and thus, to ensure the commercial profit, it needs to negotiate some bilateral contracts in advance prior to participating in the day-ahead market. From the empirical point of view, most relevant decisions made by a VPP as well as its coalition members in short-term and mid-term energy transactions involve a significant level of data uncertainty. For this reason, an efficient MILP model based on robust optimization approach is proposed to enable informed decision making under different levels of uncertainty. The flexible feature embedded in this tool with respect to solution accuracy and computational burden would be advantageous to the VPP. The efficiency and applicability of the proposed method is illustrated and analyzed through different scenarios, and thereby conclusions are drawn. (C) 2015 Elsevier Ltd. All rights reserved.