Most of the structure/activity rules emerged from the initial studies by Rosenberg and Cleare and Hoeschele have now been questioned. Specifically: (i) trans compounds are inactive, (ii) charged compounds are inactive, (iii) complexes having only one leaving group are inactive, (iv) only complexes with two amine ligands, each carrying at least one H atom, are active. Exceptions to the first of these rules will be the subject of this article. These ‘exceptions’ which frequently show activity against cisplatin resistant tumour cells, fall in four classes: (1) trans-[PtCl₂(L)(L′)] with L and/or L′=pyridine-like ligands; (2) platinum(IV) complexes of formula trans-[PtCl₂X₂(L)(L′)] with X=hydroxide or carboxylate, L=ammine, and L′=amine; (3) trans-[PtCl₂(L)(L′)] with L=alkyl-substituted amine and L′=isopropylamine; and (4) trans-[PtCl₂(L)(L′)] with L and/or L′=iminoether. Greater inertness in biological medium appears to be a common feature of these compounds. Increased binding affinity for alternating purine–pyrimidine sites and enhanced interstrand cross-linking ability was found for the first and third class of compounds. Inter-strand cross-links and single-strand breaks were both proposed as cytotoxic lesions for the platinum(IV) species which presumably require reduction to platinum(II) prior to their interaction with DNA. Finally, stable monofunctional adducts with duplex DNA, causing unique local distortions in DNA which are able to inhibit in vitro DNA and RNA synthesis, were found for platinum–iminoether complexes. The recent development of new highly active platinum based drugs that do not fit the classical structure-activity rules indicates the need for a reappraisal of these rules. It is unlikely that any new widely applicable relationship will emerge. All of which goes to show that serendipity still contributes much to the study of Pt drugs, which is only appropriate given how they began.