Edelson et al.(7) first reported the use of extracorporeal photopheresis (ECP) to treat cutaneous T-cell lymphoma (CTCL) in 1987, and since then several studies reporting response rates and survival data have appeared in the literature. Several modes of action have been proposed for ECP, In CTCL there is an accumulating body of evidence to show that 8-methoxypsoralen-treated cells display increased quantities of antigenic peptides at their cell surfaces, and this in turn leads to an enhanced cytotoxic response against the neoplastic T-cell population. This mechanism requires the presence of malignant cells in the peripheral circulation, and may account for the observation that ECP produces higher response rates in erythrodermic CTCL than at other stages of disease. However, patients with inflammatory skin diseases such as reactive erythroderma may also respond to ECP, and it is therefore crucial that a diagnosis of Sezary syndrome is confirmed by demonstrating a clonal population of T cells in the peripheral blood. Unfortunately, most: studies have not employed T-cell receptor gene analysis routinely, and this may account for the different response rates and survival data reported with ECP in the literature. To date, ECP has not been tested in a randomized study against conventional forms of therapy.