In handwashing experiments with Salmonella typhimurium the effect of chlorhexidine (CHX) on the pathogenicity of surviving bacteria was assessed with and without a neutralizer in a mouse model of infection. Without neutralizer the LD50 of CHX handwash fluids was raised. Neutralizer in suspensions of untreated bacteria caused a reduction of LD50 up to 1.2 logs. Thus, in contrast to soap or alcohol, CHX without neutralizer exerted a slight 'depathogenizing' action and neutralizer a slight 'pathogenizing' effect in the experimental model used. However, in comparison to the efficiency of handwashing procedures which reduce the number of bacteria available for transfer by at least 3.0 to 4.2 logs, the size of these effects seems to be negligibly small and unpredictable. Therefore, the single most important parameter in assessing the potency of disinfectants remains the reduction of viable counts with time.