ICOS and DIAMOND are two commercially available, semi-automated HPLC solvent optimization software packages. The resultant optimized chromatographic separation is dependent on a combination of the operator's objective, the capability of the software system and the appropriateness of the data input. The latter contains components that represent the match between the requirements of the algorithms used and the information content of the data on which those algorithms operated. Knowledge about the sample content, stability and potential sample-solvent interactions can have a significant effect on the quality of the optimal solvent composition that is calculated. The results generated during the optimization of the separation of a mixture of U-83,757 and a variety of related compounds illustrate the need to consider the significance of the contribution to the calculated optimal separation of each of these potential pitfalls, both individually and in combination with the mode of operation of the relevant algorithms. Our results indicate that the quality of the final result is highly dependent on the intelligence content of the data used.