The aim of this study was to assess the effects of reductions in the number of isolates tested by phage-typing on the recognition of outbreaks of salmonellosis. Five outbreaks (categorized as 'small', 'medium' or 'large') which occurred in England in 2005 were used as examples. The outbreaks were caused by serotypes which were subdivided by phage-typing. Results indicated that reducing the number of isolates phage-typed would have an impact on the surveillance system, with one outbreak likely to have been missed altogether. However, this does not have a great effect on the 'time-to-detection' for the other outbreaks. Assuming no testing for phage-typing was undertaken it is likely that two out of five outbreaks would not have been detected. Assessing the value of phage-type information is important not only in deciding on the efficiency of the current surveillance system but also in providing a basis upon which to assess more detailed typing methodologies such as an antibiogram of molecular profile.