Abstract
Recently a great deal of emphasis has been placed on the role of risk in theories of accident involvement. This may be exemplified by risk homeostasis theory, which argues that the level of risk people are willing to accept is the sole determining factor in overall accident involvement. The evidence for and against this position is reviewed and it is concluded that there is little evidence in favour of the theory. Several theoretical and methodological inconsistencies are noted. It is concluded that an increased knowledge of the limitations of human risk perception will prove useful in understanding how people react to human error and accident involvement.