The concept of the myofibroblast is relatively new, since this cell type was defined less than 30 years ago, and there is as yet no firm consensus as to how a myofibroblast should be defined either morphologically or phenotypically. Because its attributes place it midway between a fibroblast and a smooth muscle cell and because it appears capable of functional and phenotypic modulation, some authorities argue that there is no such thing as a truly myofibroblastic tumour. However, since most diagnostic tumour pathology is based on phenotypic features of the excised lesion, it is an unavoidable fact that a wide range of soft tissue tumours show convincing myofibroblastic features. Lesions in this category fall into four main groups--the family of reactive fasciitis-like lesions, a group of benign lesions most of which have been fairly recently recognized (e.g., mammary myofibrolastoma, intranodal myofibroblastoma, angiomyo-fibroblastoma and dermatomyofibroma), the locally aggressive fibromatoses (either superficial or deep) which share features of fibroblasts and myofibroblasts in varying degree and, finally, sarcomas showing myofibroblastic differentiation. These latter include low grade lesions such as so-called infantile fibrosarcoma, inflammatory myofibroblastic tumour and a distinctive form of low grade myofibroblastic sarcoma, as well as some of the high grade lesions formerly known as MFH. Based on the advances made in the past 20 years, it makes no sense to deny the existence of myofibroblastic tumours (especially in an era when histogenetic concepts have crumbled)--instead we need to work towards a reproducible and agreed definition of the myofibroblast.