This work aims at proposing a sliding mode observer (SMO) based nonlinear multivariable sliding mode controller (SMC) and globally linearizing controller (GLC) for a proton exchange membrane fuel cell (PEMFC). First, a nonlinear dynamic model of a methanol reformer is developed to produce hydrogen for the PEMFC. The reformer model has been verified and then its parameter values are modified to scale-up the reformer for fitting with a real time fuel cell. Subsequently, a nonlinear PEMFC model is formulated and validated with experimental data. The SMC consists of a controller and an estimator, whereas the GLC consists of a transformer, an estimator and a dual-loop external proportional integral (PI) controller. A transformer that relates the manipulated variable with the external control output is developed using the differential geometry. For both the control schemes, a nonlinear sliding mode observer is formulated to avoid any additional requirement of sensor to measure other than the controlled variables. Chattering occurred in the SMO and SMC due to very large frequency stroke of the sign function to minimize the estimation error has been eliminated by using a saturation function. Finally, the proposed SMO based SMC and GLC structure are tested for the integrated methanol reformer-PEMFC system. Observing an excellent estimation performance of the SMO, we investigate a comparative performance between the SMC and GLC with reference to a dual-loop PI controller. (C) 2017 Elsevier Ltd. All rights reserved.