Circulating fluidized beds (CFBs) are used widely in the chemical industry. Knowing or estimating the bed height in the standpipe and the solids circulation rate are essential for effective control of the system. This paper incorporates a 2-region model to calculate the bed height in the standpipe with a Kalman filter algorithm to estimate the solids circulation rate (SCR). Simulations of both the standpipe bed height and SCR were compared with experimental data and shown to give good agreement. In addition, a neural network method was applied to model the entire cold flow CFB system and measured data sets were used to train the neurons of the network. Finally, a linear controller was applied to control both the bed height and solids circulation rate to desired set points. Simulations were performed for both positive and negative step inputs for both variables and satisfactory control was demonstrated using this controller in combination with the neutral network and Kalman estimator. Published by Elsevier B.V.