A stochastic model representing annular flow in a tubular reactor is proposed. Numerical simulation was utilized to generate sample paths fitting the residence time distributions (RTD) of the system. The model was constructed from basic diffusion equations with the additional consideration of random effects disturbing the system, thus yielding a stochastic partial differential equation. The stochastic model is simulated using the EulerMaruyama procedure. Experimental data from three tubular polymerization reactors were well fitted by the model. The model encompassed the two deterministic parameters, mean residence time and Peclet number, as well as three stochastic parameters: stochastic relevance (b), updated time (Delta T) and the seed that begins the Wiener process. The satisfactory results indicate that the model constitutes an important step toward comprehending the complex fluid dynamics of tubular flow systems. (C) 2010 Elsevier Ltd. All rights reserved.