This paper differs markedly from the others that are being presented at this symposium, in that it does not describe a particular technique, or have a single unifying theme. Its aim is to draw to the attention of the purging community a list of variables that we all have tended to regard as only secondarily important to our primary goal of target cell depletion, or collection. It is often difficult to remember that ex vivo purging treatments are composed of multiple steps, and that, in turn, purging is a single component of a complex treatment protocol. In order to cope with this diversity, the natural tendency has been to focus on manageable segments. While this can speed development, we need to constantly be aware that there are multiple factors, both within and outside of our own particular segment, that can impact on the final result. Variables that we often regard as unimportant, or techniques that are thought of as routine, may ultimately be influencing clinical outcome. It would be naive to suggest that we can ever have control over all of the variables in any procedure that is part of a clinical treatment. One purpose of this paper is simply to draw the existence of these factors to the attention of investigators, and suggest that their potential impact may not always be fully appreciated. Secondly, the selective separation of cells from bone marrow is sufficiently well established that it should now be possible to evaluate the technology, reagents and variables in a systematic and collaborative fashion. While we may not emerge from such an exercise with unity of option, we may at least lay the foundations for the further development of this type of therapy, and provide the framework for addressing future questions.