Hedgehog (HH) signalling is involved in the development of numerous embryonic tissues. In humans,germline mutations in hedgehog pathway components cause congenital malformations and somatic mutations are associated with cancers. The basic framework of the HH pathway was elucidated in the fruitfly, Drosophila melanogaster, and this pathway is largely conserved in vertebrates, although some important differences have been noted. The current paradigm of the "canonical" pathway views HH signalling as a series of repressive interactions which culminates in GLI-mediated transcriptional regulation of a variety of cellular processes. Definitions of "non-canonical" signalling stem from examples where the response to HH morphogen deviates from this paradigm and, according to current reports, three general scenarios of noncanonical HH signalling can be defined: (1) Signalling that involves HH pathway components but which is independent of GLI-mediated transcription; (2) Direct interaction of HH signalling components with components of other molecular pathways; and (3) "Non-contiguous" or "atypical" interaction of core HH pathway components with one another. Currently, the evidence supporting non-canonical HH signalling is not conclusive. However, Sonic hedgehog (SHH) has been shown to regulate cell migration and axon guidance in several contexts, and some of these processes are independent of downstream components of the HH pathway, and presumably the transcriptional response to morphogen. Furthermore, biochemical studies have shown that the HH receptor, PTCH1, can directly interact both with Cyclin B1 and caspases, to inhibit cell proliferation and to promote apoptosis, respectively, and that these functions are inhibited in the presence of morphogen. Genetic analysis of orthologues of the HH pathway in nematode worms further supports the notion that PTCH1-related molecules can function independently of other components of the canonical HH pathway, and the phenotypes of mice with point mutations in the Ptch1 gene offer clues as to the processes that non-canonical HH signalling might regulate. While none of these evidences are conclusive,collectively they point to the existence of added complexity in the HH pathway in the form of non-canonical pathways. A major difficulty in studying this problem is that canonical and non-canonical pathways are likely to act in parallel, and so in many situations it will not be possible to implicate non-canonical responses in certain cellular processes simply by excluding a role for the canonical pathway-directed analyses of non-canonical HH signalling are therefore necessary. The aim of this review is to present the cumulative evidence supporting non-canonical HH signalling, with the hope of promoting further enquiry into this area.