A new case of Sweet's syndrome (acute febrile neutrophilic dermatosis) associated with a malignant hemopathy is presented. The blood disease was a chronic myelomonocytic dysmyelopoiesis which was discovered during the eruption and resulted in the patient's death within a few months, probably through acutization. The skin lesions were atypical, bullous and ulcerated. On this occasion, the international literature concerning all cases of Sweet's syndrome associated with malignant or premalignant hemopathies is reviewed. Several concepts emerge from this study: the association is frequent (about 20 p. 100 of all published cases of Sweet's syndrome); there is a strong predominance of granulocytic hemopathies over lymphoplasmocytic and monocytic hemopathies; the blood disease is revealed by the skin eruption in some 50 p. 100 of the patients; there are frequent chronological relations between Sweet's syndrome and the events that occur in the course of the hemopathy; finally, the association is usually of poor prognosis. A comparison with Sweet's syndrome unassociated with a blood disease showed only three significant points: the frequency of bullous lesions, of the initial anaemia (the most important element) and of extreme figures in leucocyte counts (leucopenia or major hyperleukocytosis). The atypical character of the skin lesions in the patient presented here incites to discuss the nosological relationship between Sweet's syndrome and bullous pyoderma, an entity closely associated with hemopathies. It has recently been suggested by several authors that this anatomico-clinical kinship should be turned into a wide spectrum of acute neutrophilic dermatoses, with typical Sweet's syndrome at one end and Pyoderma gangrenosum at the other end. The interface between this spectrum and haemopathies seems to be maximum at its intermediate stage: the bullous and superficially ulcerated lesions. The aetiology and pathogenesis of this new nosological entity are uncertain. The presence of chemoattractants or of polymorphonuclear cell abnormalities is still open to discussion. The relationship between the entity and leukocytoclastic vasculitis has recently been questioned.