Reverse osmosis (RO) membranes play a key role in wastewater treatment units as they are used to remove salts and other pollutants effectively. RO membrane performance is affected by many different factors such as feed characteristics and operational parameters during operation. The aim of this study is to analyse the influence of feed characteristics of municipal wastewater (conductivity, oxidation reduction potential (ORP), total suspended solids (TSS), turbidity and chemical oxygen demand (COD)) and operational parameters (feed pressure, flow rate and temperature) on RO membrane performance in a municipal wastewater recovery plant using machine learning (ML) techniques. XGBoost, random forest, artificial neural networks (ANNs) and multiple linear regression (MLR) were employed to predict three RO membrane performance indicators (pressure difference across membranes, salt passage and permeate flow rate). The methods that can predict salt passage, permeate flow rate and pressure difference among membranes with the highest accuracy were found as ANNs, random forest and MLR, respectively. Considering the developed models, temperature was found to be the variable affecting all three RO performance parameters. Salt passage was found to be highly affected by feed water conductivity and feed flow rate was determined to be the most influential parameter for the permeate flow rate and pressure difference.