Based on the hypothesis of Goldie and Coldman, rapid cyclic alternating chemotherapy has been supposed to be a favorable treatment modality in small cell lung cancer. This approach has been tested in a large series of trials since the late 70's. A few trials performed randomization between alternating and continuous treatment after a period of initial common continuous therapy. The results of these studies are controversial and their interpretation is complicated by the effects of the continuous pretreatment. Recently, most trials were designed as 2-arm approaches with a comparison of continuous standard therapy based on the CAV- or CMC-protocol with an alternating schedule often consisting of the CAV- or CMC-derived combinations and a second regimen including cisplatinum and/or etoposide. In these trials the addition of the second regimen in the alternating treatment arms has improved the treatment results. However, these trials could not clarify whether this advantage was due to the concept of alternating treatment or to the high activity of new drugs given early in the course of therapy in the alternating treatment arms. To ameliorate this weakness, 3 studies were designed as 3-arms approaches using both alternating protocols as continuous control arms. Whereas one investigation noticed an advantage of the alternating schedule, the two others observed no difference. These studies did not report their criteria for shifting non-responding patients from continuous treatment to second line therapy. The longer these patients stayed on the continuous protocol the more the study design favoured the alternating therapy.(ABSTRACT TRUNCATED AT 250 WORDS)