The entire issue of unconscious influences on thought and behaviour has been reinstated as a matter of considerable interest to psychologists of varying persuasions. Investigators interested in perception have recently (and once again) demonstrated that people can perceive and be influenced by information that is not well represented in consciousness (i.e., is not noticed). Research in social psychology, on the other hand, seems to have shown that even when information is noticed, it can affect a person''s thought and behaviour in a manner that is not appreciated or understood. Such uncomprehended influences are properly characterized as having an unconscious impact on thought and behaviour. Although such empirical research helps to illustrate the importance of unconscious influences, my thesis actually follows from two a priori assumptions that seem to me compelling: (a) introspection has no privileged access to the causal connection linking thought and behaviour to their determinants, and (b) observation (of one''s own behaviour, for example) never entails a particular account of the phenomena observed. Consequently, some theory, however informal or implicit it may be, is required to account for one''s thought or behaviour. There is no reason simply to assume the validity of such a theory, however, and insofar as it is invalid - that is, insofar as it overlooks or disavows relevant considerations - the thought or action can be described as having been unconsciously influenced. Some implications of this view for the human condition are presented in a final section.