A quality assurance programme was used to evaluate community and primary care based preschool surveillance using the National Child Health Computer System in 40 examination centres. Quarterly reports were generated from returns from clinical medical officers and general practitioners to list non-attenders, uptake, and timeliness for the four preschool checks. These provided rapid and comparative feedback on personal performance for participating health professionals and led to marked rises in recorded timeliness and uptake against preset targets. Pre-existing uptake was highest at the 6 week check with least overall improvement. Greatest improvements occurred at the 18 month health visitor check but, in general, results plateaued when the programme had been in use for 12 to 18 months. Particular problems such as data legibility and mobile populations were identified and solutions formulated. It is postulated that improvements in performance were due to enhanced professional motivation as no other factors changed. This system provides a valuable contribution in the light of changing patterns of service provision.