The aim of this note is to show that the algorithm recently proposed by J. Bard (1983) is not guaranteed to converge to the desired solution of a Linear Two-Stage Optimization Problem (LTSOP), unless additional regularity conditions are imposed.