The aim of this study is to assess the effects of fulvic acid and iron oxide on the Fe2+ and Mn2+ removal by an aerated-submerged membrane system. To construct empirical models for the relationship, polynomial regression method is applied. Ultrafiltration hollow fiber membrane module is submerged into aeration tank. The pressure changes in the membrane system are continuously monitored by means of pressure gauge. Dissolved organic carbon, UV254, iron and manganese concentrations are measured in the raw, unfiltered and filtered waters. While iron oxide's presence increases the removal efficiency of Fe2+ and Mn2+, high concentration of fulvic acid affects adversely. The relationship between these contaminants and membrane fouling is investigated. Also, in order to model the relationship, regression analysis using least squares (LS) approximation method is performed. This method is based on a polynomial of degree n to compute the best approximation to data. Mathematical models are evaluated and compared to identify an adequate model to represent the flux data collected with different ingredients. The results obtained by LS show that the method is feasible.