Background: Continuity of care is a central objective of community psychiat
ric services, but there is no consensus about its measurement. Aims: We dev
eloped measures of continuity of care suitable for routine use, and measure
d continuity and individual patient outcome over a period in which communit
y services were developing. Method: One hundred patients with severe mental
illness receiving continuing care from two sectorised services were sample
d and interviewed. Data were collected concerning their care over 20 months
prior to interview. After 20 months prospective followup, they were rr-int
erviewed. Continuity was defined as: perceived accessibility of services an
d knowledge about them, the number of keyworkers in a defined period of tim
e. and the proportion of time out of contact with services. Results: Contin
uity of care improved significantly on all measures over the period of the
study. Individual patient outcome also improved, but in multiple regression
models including clinical and demographic variables. measures of continuit
y were not significant predictors of outcome. Continuity was similar for wh
ite and non-white patients. Conclusion: Simple measures of continuity are u
seful in evaluating changes in the process of care, but they are not straig
htforwardly related to individual outcome.