The international reference strains of variola major (Harvey) and of variola minor (Butler) were grown in cultures of skin and muscle cells from human embryos. The development of infective virus, complement-fixing antigen, haemagglutinin and cytological changes were followed at four temperatures between 35 and 40 degrees C. No significant difference was found in the amount of virus produced by Harvey or Butler viruses at any of the experimental temperatures, but Harvey attained the plateau titre at 16 h, some 4 h ahead of Butler in the cultures incubated at 38 degrees C. Harvey also produced a higher and more prolonged yield of virus in the extracellular medium of cultures, inoculated at low multiplicity and incubated at 37 degrees C. At 38 degrees C small inocula of Harvey produced foci which developed and spread till the whole culture was necrotic; Butler foci did not spread and remained relatively undeveloped at this temperature.Staining with acridine orange showed the development of cytoplasmic inclusions at all temperatures up to 39.5 degrees C, at which temperature most inclusions remained round and discrete and susceptible to digestion with DNase. The yield of virus at all temperatures correlated well with the number of cells in which the DNA cytoplasmic inclusions became irregular in outline, diffuse and insusceptible to digestion with DNase. It was concluded that elevated temperatures, up to 39.5 degrees C affected principally a maturation phase in the development of the virus.Equal amounts of complement-fixing antigen were produced at all temperatures and by either virus, but the late product, haemagglutinin was depressed at elevated temperatures much more in cultures infected with Butler than in cultures infected with Harvey. This was clearly shown by haemadsorption; in cultures, infected at high multiplicity and incubated at 39.5 degrees C Harvey gave semi-confluent haemadsorption, while only an occasional haemadsorbing cell could be found in the cultures infected with Butler virus.It is concluded that there were two ways in which temperature affected the growth of variola viruses in cultures of human embryo skin and muscle cells. The total yield of virus was reduced by inhibition of a stage in the maturation of the virus with a cut off point between 39.5 and 40 degrees C at which no cells produced infective virus, and with little observable differences between Harvey and Butler viruses. Changes at the surface of virus-infected cell, involving virus release and haemagglutinin, were affected independently of virus maturation; in these changes Butler was much more sensitive than Harvey to elevated temperatures. The relevance of these observations to the development of variola major and variola minor in man is discussed.