The idea that child witnesses are unreliable because of their high suggestibility is a common one. However, it is questionable if suggestibility can be attributed to individuals without considering the situational determinants of the phenomenon. This paper reviews studies of suggestibility in adults and children with particular reference to its determinants. The possibility that interview techniques might be developed which could enhance the reliability of child witnesses is considered.Within an adversarial system of justice the credibility of any witness testifying in a court of law is liable to be challenged during cross-examination. However, the credibility of a witness who is also a child faces additional challenges. One such is the belief expressed by many adults that children's testimony can readily be altered by what is said to them or by what questioners appear to require of them (cf. Ross, 1908; Lipmann, 1911;Andrews, 1964;Yarmey and Jones, 1983;Baxter and Davies, 1986). In other words, many adults, and hence many potential jurors, seem to believe that a major problem with child witnesses is their relatively high suggestibility.But what is the nature of suggestibility, and of children's suggestibility in particular? If it is a stable feature of childhood, analogous to a personality trait, then children's testimony should perhaps be distrusted as a matter of course. It may be that children are especially likely passively to absorb any information which appears to concern an event which they have witnessed, such that they will subsequently incorporate that information into their accounts of the event. This was apparently the view of Otis (1924), who argued that suggestibility is a response which:follows the stimulus directly, the subject being in a way an automaton. With adults the use of the higher centres is in abeyance while with children the development and use of the higher centres is not complete (p. 9).On the other hand, it may be that a main problem with child witnesses is that they are vulnerable to a much broader range of social pressures than are adults, such that this kind of pressure plays a disproportionately large role in determining children's responses to questioning. If this is the case then it may be that the reliability of children's testimony could be substantially enhanced by ensuring that , at least during formal questioning, social pressures on children to answer questions in particular ways are identified and minimized.