This paper aims to investigate the impact of different Electric Vehicle (EV) penetration on quasi real-time Volt-VAR Optimization (VVO) of smart distribution networks. Recent WO solutions enable capturing data from Advanced Metering Infrastructure (AMI) in quasi real-time to minimize distribution networks loss costs and perform Conservation Voltage Reduction (CVR) to save energy. The emergence of EVs throughout distribution feeder increases grid complexity and uncertainty levels that could affect AMI-based VVO objectives. Hence, this paper primarily introduces an AMI-based WO engine, able to minimize grid loss and Volt-VAR control assets operating costs while maximizing CVR benefit. It then presents a real-time co-simulation platform comprised of the VVO engine, grid model in a real-time simulator and monitoring platform, communicating with each other through DNP.3 protocol, to test the precision and performance of AMI-based WO in presence of different EV penetration levels. Accordingly, 33-node distribution feeder is studied through different EV penetration scenarios. The results show significant changes in AMI-based WO performance especially in CVR sub-part of WO according to EV model and type. Thus, this study could lead near future WO solutions to gain higher levels of accuracy and efficiency considering smart microgrid components such as EV in their models. (C) 2016 Elsevier Ltd. All rights reserved.