Oxazaphosphorine cytostatics (OX) such as cyclophosphamide (CP) and ifosfamide (IF) are the most effective alkylating cytostatics. All efforts to understand and improve the cause of their good antineoplastic effects have so far been unsuccessful. The cause of this failure was the fact that the results of in vitro tests were uncritically transferred to in vivo conditions. In doing so the fact that the metabolite spectrum of OX in vivo is different from that in vitro was overlooked. In vitro, the pharmacologically active metabolite (OX-ALD) is converted into the ultimate alkylating metabolite OX-phosphoreamide mustard (OX-PAM) by β-elimination of acrolein. In vivo, however, no acrolein is formed because OX-ALD is cleaved enzymatically by phosphodiesterases into OX-PAM and the pro-apoptotic aldehyde 3-hydroxypropanal (HPA), which - and this is special for OX – boosts the p53-dependent apoptosis caused by DNA alkylation. Based on the discovery of HPA as a CP metabolite, a mechanism of action for OX was postulated on the basis of which OX-ALD-perhydrothiazine and OX-ALD-thiazolidne derivatives with different alkylating functions were synthesized. The thiazolidines and perhydrothiazines are OX that spontaneously hydrolyze to OX-ALD. Therapy experiments in CD2F1 mice bearing advanced, solid growing P388 tumors showed that the OX-ALD-derivatives, in which a chloro-ethyl group which makes easily repairable DNA interstrand crosslinks is substituted by a mesyl-ethyl group which makes poorly repairable intrastrand crosslinks, are orders of magnitude more effective. Based on these findings, mesyl-I-aldophosphamide-perhydrothiazine (SUM-IAP) was developed as a model substance for a new generation of OX and tested experimentally. The development of a scheme for the mechanism of action of OX and its validation in experiments with SUM-IAP is reviewed.