The accuracy of the field diagnosis is of crucial importance in the successful implementation of a screening programme. In Sri Lanka, the feasibility of using primary health care workers for oral cancer and precancer screening was pilot tested between 1980 and 1982. This paper analyses in detail the categories of false positives referred and the false negatives detected following screening. Among 660 subjects who arrived at the referral centre following screening, 64 (10 per cent) had no apparent oral lesion. The predictive value determined in this study was 58 per cent for a referrable lesion and 98 per cent for non referrable cases.