There is a need to develop and test methods of audit of the consultati
on process in occupational medicine and to draw conclusions from the f
indings. An external audit was carried out on 313 randomly selected co
nsultation records from the occupational health services of three heal
th boards over a 26-month period. Additionally, nine months after the
date of the first externally audited sample, a prospective internal au
dit was commenced within one board. In 58 of the consultations selecte
d (19 per cent), audit was not possible mainly because records could n
ot be found. In an appreciable proportion of the rest, there were shor
tcomings in the referral record or in the consultation record. The con
sultation records of 'career' occupational physicians were better than
those of 'non-career' physicians, especially in recording specific di
agnoses (91 per cent vs. 67 per cent; P < 0.0001) and in conclusions r
egarding occupational implications (86 per cent vs. 74 per cent; P < 0
.0001). Multivariate regressions showed a highly significant improveme
nt associated with prospective internal audit in one item, namely the
record of occupational implications. However, the magnitude of this di
fference before and after audit was smaller than the independent diffe
rence arising from the career status of the physicians. The value of m
edical audit in improving quality of care and education in occupationa
l medicine is discussed.