Biodiesel is a nontoxic, biodegradable, renewable diesel fuel obtained from lipid feedstocks. The most common production process for biodiesel is through the batch transesterification of vegetable oils with methanol. In the batch process, excess catalyst which is used to drive the reaction to completion can result in high material and processing costs and the degradation of components found in lipid feedstocks which impart color to the resulting biodiesel. Much work on heterogeneous catalysts has been done in the past decade; however, most of the work was done at temperatures and pressures well above those used in atmospheric batch processes. Under these conditions, the reaction rates of homogeneously catalyzed reactions would be enhanced. However, equilibrium limitations do require substantial amounts of catalyst to be present to drive the reactions to completion. Reducing the amount of homogeneous catalysts required for the transesterification would reduce the need for washing of the fatty acid methyl esters (FAME) product and the formation of soap in the biodiesel and improve the color of the resulting glycerol and in many cases the resulting biodiesel. In the present study, the transesterification of canola oil with methanol was investigated at varying catalyst concentrations and residence times (RT) in a continuous membrane reactor. Prior to all experiments, the free fatty acid in the canola oil was neutralized with sodium hydroxide. Experiments were performed at 0.0, 0.01, 0.03, 0.05, 0.1, 0.5, and 1 wt % sodium hydroxide on an oil basis. The methanol:oil mole ratio used in the experiments was 24: 1. It was found that a base catalyst concentration above 0.05 wt % for a I h residence time (RT) and above 0.03% for a 2 It RT resulted in the steady-state biodiesel production via the membrane reactor. Such catalyst amounts are 10-33 times lower than those employed in the industrial production of biodiesel (e.g., 0.5-1 wt % sodium hydroxide concentration). Mono- and triglycerides were not detected in any of the biodiesel produced in this single reaction step process. The RT in the reactor displayed no obvious effect on permeate composition. The results demonstrated that maintaining the phase separation between the oil and methanol at all times during the reaction and the ability of the membrane to separate these phases were key in obtaining high quality product. Unreacted triglycerides and unsaponifiable, oleophilic impurities in the lipid feedstock were not integrated into the final biodiesel as in conventional processes due to the presence of two phases in the reactor. The results illustrate the advantages in using a membrane reactor to produce biodiesel, where products are continuously removed from the reactor in a different phase than the lipid reactant.