Estimation of the ethanol content in flexible-fuel vehicles (FFVs) has attracted increasing attention given the possibility of a variety of fuel blends, beyond E10 and E85, at the pumping stations. In this paper we develop an ethanol estimation algorithm for FFVs equipped with cylinder pressure sensing and direct injection. The algorithm utilizes the cylinder pressure measurements to infer the charge cooling effect associated to evaporating fuel with various ethanol concentrations [Oliverio et al. (2009)]. The ethanol detection feature using the charge cooling effect is named LHV estimation due to the fact that it relies on the Latent Heat of Vaporization characteristics of the fuel. The LHV estimation is shown to be more robust to fuel injector drifts than the traditional ethanol content estimation based on the stoichiometric air-to-fuel ratio (SAFR) identified through the closed-loop air/fuel regulation using the exhaust gas oxygen (EGO) sensor [Ahn et al. (2010)]. In this paper, a method is presented to enable robust ethanol estimation in the presence of fuel injector drifts by integrating the LHV and SAFR based detection features. Dynamometer testing and simulations of a 2.0L turbocharged spark ignition direct injection (SIDI) engine has demonstrated the feasibility of implementing the proposed method at specific operation conditions.