During an outbreak of acute Salmonella braenderup gastroenteritis we performed a standardised interview encompassing questions on clinical symptoms in 156 (127 adults and 29 children) of 215 identified patients. Sequential stool cultures were obtained for up to five months in these 156 cases. We restricted the analysis to the 122 patients with at least 3 or more available cultures. They were treated with a fluoroquinolone, trimethoprim-sulfamethoxazole, or not treated with antibiotics, according to the decision of the practitioners. For this reason, a randomised double blind study was not possible. Minimum inhibitory concentrations (MIC) of the prescribed drugs were measured for representative isolates before and after treatment. The most frequent symptoms were diarrhoea (98%) and abdominal pain (96%). Vomiting occurred in 43% of cases. Children were more severely ill. Seven weeks after acute gastroenteritis, stool cultures were still positive for salmonella in 71% of the 22 children and 30% of the 100 adults examined (p < 0.002). This rate decreased progressively in both groups to 5 and 3% respectively at 20 weeks (n.s.). Among adults, no significant difference in enteric carriage over time could be demonstrated between untreated patients and those treated with either a fluoroquinolone or trimethoprim-sulfamethoxazole. MIC for salmonella isolates remained unchanged after treatment. In a cohort of patients infected with a single strain of salmonella, fluoroquinolone therapy of acute gastroenteritis failed to influence the duration of enteric carriage, despite continuing susceptibility of the strain. In children, the rate of clearance of Salmonella braenderup from stool was statistically lower until the tenth week after the acute disease, but there was no further difference after 5 months.