in four months of observation at stafford house, an attempt was made to discover whether or not residents were receiving "humanized" care and if so, if this humanization was a result of, or was enhanced by, the implementation ofthe primary nursing method. after a tortuous analysis of field notes, what became clear was that there was no obvious pattern of humanization. although there were numerous examples ofhumanizing resident care (i.e. shared decision making, freedom of action, etc.) there were just as many instances of dehumanization. the lack ofa clear pattern was disturbing and frustrating. ifprimary nursing was a structural attempt to humanize patient care, and if it was being implemented by all nursing staff, and with all residents, why wasn’t there a clear pattern emerging? instead, repeated analyses of field notes revealed a seeming randomness of humanization. this finding may appear to indicate that primary nursing is not an innovation which humanizes health care. that is not necessarily the case. this thesis puts forth the argument that theoretically primary nursing can be a humanizing intervention within the parameters outlined by howard (1975). the reason for its lack of success at stafford house would appear to be a result ofa number of factors. external constraints (ie inadequate funding, government regulation etc.) were combined with a faulty, and unrealistic interpretation of what primary nursing is and how it can be operationalized. individual actors are forced to act within an organizationally and individually imposed structure which is itself dehumanizing.