In 1992 it was decided to re-evaluate the Widal slide agglutination test as a rapid diagnostic test for typhoid fever in Papua New Guinea. This was in response to an apparent increase in the number of false positive Widal slide agglutinations occurring using an O cut-off titre greater than or equal to 40 which was previously shown to be appropriate in 1987. The results of the re-evaluation indicated that the Widal test using a diagnostic cut-off titre of ≥40 lacked specificity and was no longer appropriate for this population. A new O antibody titre of ≥160 was recommended as a diagnostic titre for typhoid fever in PNG. The fall in the specificity of the Widal slide agglutination test over the five-year period between the initial assessment and the re-evaluation is due to an increase in general population antibody levels caused by the changing pattern of typhoid in the community. Before 1987 typhoid fever occurred as sporadic, isolated outbreaks and most people living in the highlands of PNG were immunologically naive. By 1992 typhoid fever had become a well-established endemic disease and many more people had been exposed to Salmonella typhi and as a result developed antibodies. We have been able to demonstrate clearly a remarkable change in the immune status of the community, in which the proportion of healthy individuals with a Widal tube O agglutination titre of 40 or more rose from 0 to 56% in the short span of five years. Factors influencing the usefulness of the Widal agglutination test in a population include the endemicity of typhoid fever, the immune response to typhoid, general population antibody levels to Salmonella typhi and the ability to select an appropriate diagnostic cut-off titre. Though its usefulness will vary with time and place, the Widal agglutination test remains a useful diagnostic test under appropriate circumstances.