The persisting ancient view of cancer as a contagious disease ended with 19th century scientific investigations which seemed to show it was not. The resulting dogma against an infectious cause for cancer produced great prejudice in the scientific community against the first report of an oncogenic virus by Rous early in the 20th century and, even in the 1950s, against Gross's finding of a murine leukaemia virus and a murine virus causing solid tumours. The Lucké frog renal carcinoma virus was the first cancer–associated herpesvirus. Intriguingly, an environmental factor, ambient temperature, determines virus genome expression in the poikilothermic frog cells. Although an α–herpesvirus, Marek's disease virus of chickens shares some aspects of biological behaviour with Epstein–Barr virus (EBV) of man. Very significantly, its lymphomas are the first naturally occurring malignancy to be controlled by an antiviral vaccine, with implications for human virus–associated cancers. The circumstances and climate of opinion in which successive γ–herpesviruses were discovered are described. The identification of EBV involved two unconventionalities: its finding in cultured Burkitt's lymphoma cells when no human lymphoid cell had ever been maintained in vitro , and its recognition in the absence of biological activity by the then new technique of electron microscopy. These factors engendered hostility to its acceptance as a new human tumour–associated virus. The EBV–like agents of Old World apes and monkeys and the T–lymphotropic γ–herpesviruses of New World monkeys were found at about the same time, not long after the discovery of EBV. For many years these were thought to be the only γ–herpesviruses of non–human primates; however, very recently B–lymphotropic EBV–like agents have been identified in New World species as well. Mouse herpesvirus 68 came to light by chance during a search for arboviruses and has become important as a laboratory model because of its close genetic relatedness to EBV and its comparable biological behaviour. The discovery of Kaposi's sarcoma–associated herpesvirus six years ago was made using unconventional new methods, but, unlike with EBV 30 years before, this did not hinder its acceptance. This contrast is discussed in the context of the great progress in human tumour virology which has been made in recent years.