This paper is an investigation of the effects of a buffer stocks-supported price stabilization scheme on the Bangladesh jute sector. An econometric model of the Bangladesh jute sector is used for controlled policy experiments. A first order stochastic process is used to generate empirical probability distributions of the endogenous variables. A complete price stabilization policy, in which farm price is contained within a 7.5% band around the expected price is simulated. It is found that a substantial reduction in price instability can be achieved. While the estimated cost of the scheme is reasonable, the 95% confidence intervals are very large. The results of the simulations are also found to be highly sensitive to the assumed value of the export price elasticity.