This paper presents two new methods for the maximum power point tracking (MPPT) control of a pressure retarded osmosis (PRO) salinity power plant, including mass feedback control (MFC) and fuzzy logic control (FLC). First, a brief overview of perturb & observe (P820) and incremental mass resistance (IMR) control is given as those two methods have already demonstrated their merit in good control performance. Then, two new methods employing variable-step strategy, MFC and FLC, are proposed to address the trade-off relationship between rise-time and oscillation of P820 and IMR Genetic algorithm (GA) is used for finding the optimum parameters of membership functions of FLC. From the case-study of start-up of the PRO adopting MPPT control, MFC and FLC have shown faster convergence to the target performance without oscillation compared with P820 and IMR. These four MPPT techniques are further evaluated in case-studies of state transitions of the PRO due to operational fluctuations. It is proven that the MPPT using FLC and modified MFC has better performance than the other two methods. Finally, the paper reports a comparison of major characteristics of the four MPPT methods, which could be considered as guidance for selecting a MPPT technique for the PRO in practice. (C) 2016 Published by Elsevier B.V.