Typhoid fever caused by Salmonella typhi remains endemic to many parts of South Africa, including Natal and KwaZulu, Northern Transvaal and the Transkei. Until recently, the majority of S. typhi isolates from South Africa have remained susceptible to ampicillin/amoxycillin and chloramphenicol, and only three cases of typhoid due to multi-antibiotic resistant strains of S. typhi have been documented. Ampicillin/amoxycillin and chloramphenicol are, therefore, still recommended as first line therapy for patients with typhoid fever in this country. We describe a cluster of six cases of typhoid caused by S. typhi that was resistant to ampicillin, chloramphenicol and trimethoprim-sulphamethoxazole. All these patients presented over a 3-month period; the patients were from three adjacent districts in the Northern Natal area of South Africa. The high rate of intestinal perforation (two of six) was a direct consequence of inappropriate antibiotic treatment. Failure of surgical intervention, renal impairment as well as delay in starting appropriate antibiotic treatment were factors contributing to the high mortality (three of six). The good clinical outcome in the remaining three patients probably resulted from treatment with appropriate antibiotics; however, mild disease in two of these patients may have been a contributing factor. All isolates showed high minimal inhibitory concentrations (MIC) of greater than or equal to 256 micrograms/ml to ampicillin, chloramphenicol and trimethoprim-sulphamethoxazole. The isolates were all highly sensitive to the third generation cephalosporins (MIC less than or equal to 0.06 micrograms/ml) and quinolones (MIC less than or equal to 0.03 micrograms/ml). Conjugation studies suggest a genetic transfer of resistance, probably plasmid mediated. The presence of beta-lactamase and chloramphenicol acetyl transferase enzymes in all six isolates tested would account for the resistance to ampicillin and chloramphenicol respectively. The transfer of such plasmids to erstwhile sensitive strains could conceivably occur in this typhoid-endemic area, where sanitary conditions are poor and living conditions crowded, thus further exacerbating the problem. It is recommended that in areas where such multiresistant strains are encountered, the third generation cephalosporins or quinolones be used as empiric therapy for typhoid fever.