Puccinia recondita f. sp. tritici pathotype (pt) 64-(6),(7),(10),11 was first detected in Australia in northern New South Wales in 1990. Greenhouse tests of pathogenicity on wheat and triticale genotypes indicated that this pathotype was similar to two closely related pathotypes, but differed from them in at least three pathogenic features indicating that it had not arisen by a simple mutational event. Pathotype 64-(6),(7),(10),11 combined several pathogenic and isozymic features which, prior to its detection, were known only in two different groups of pathotypes, strongly suggesting that it had arisen via somatic hybridisation between isolates from each group. This hypothesis was supported by the RAPD phenotypes of isolates representing pathotypes present in Australia at the time pt 64-(6),(7),(10),11 was first detected. The hybrid pathotype combined virulence for Lr1 with partial virulence for gene Lr13 and was fully virulent in greenhouse adult plant tests on two hybrid wheats heterozygous for these genes. This is the first case where compelling evidence has been obtained for somatic hybridization in Prt and only the second of inter pathotype somatic hybridization in a wheat rust under field conditions.