Background. Research into the health of elderly people has found probl
ems unknown to their general practitioners. It was anticipated that an
nual checks, as required by the 1990 general practitioner contract, wo
uld help to detect these problems, yet the value of these checks has b
een questioned. Aim. To investigate the problems found by general prac
tice contractual annual checks of the over-75s and the consequent acti
ons taken; to identify patient, demographic or practice characteristic
s associated with the discovery of problems. Method. In 40 practices,
information was collected on patients over 75 years of age receiving a
health check during a 3-month period. Practices used their normal met
hods of recruitment and assessment. Practice staff were inter viewed t
o find how assessments were organized. Results. Practices saw a mean o
f 12% of their over-75s during the study; 44% were found to have at le
ast one problem. Action was taken to help resolve problems in 82% of p
atients with a problem. The most prevalent problems related to physica
l condition, and fewer functional problems than expected were found. T
here were large differences between practices in the proportions of el
derly patients seen for a check and the proportion found to have probl
ems; these were not attributable to practice size or demography. Multi
variate analysis showed that practice or patient characteristics were
poor predictors of finding problems. Conclusion. The argument in favou
r of conducting annual checks is supported by the finding that nearly
half the patients assessed were found to have problems for which some
action was taken. Some practices could increase their rate of uptake b
y modifying the organization of invitations for checks. More problems
may be found by adopting a more functionally based assessment.