The modem foundations of flexibility analysis were developed in the 1980s by Grossmann and co-workers. They formulated the approaches to solve the main problems of flexibility analysis, namely, the feasibility test, flexibility index, and two-stage optimization problem (TSOP). All these formulations are based on the supposition that during the operation stage there is enough plant data to accurately determine all uncertain parameter values. However, in practice, this assumption is likely to be untrue. To deal with this problem, the feasibility test and TSOP were extended to account for the possibility of accurately estimating only some of the uncertain parameters. However, these extensions have a drawback. They do not take into account all the data that are available at the operation stage. A formulation is discussed that will remove this drawback of the TSOP. The formulation requires working out a modification of the split and bound approach, which had been developed for solving the TSOP. Then, two computation experiments are performed to demonstrate the advantage of taking the additional plant data from the operation stage into account. (c) 2006 American Institute of Chemical Engineers.