This study examines the causes for the therapeutic failure of trimethoprim-sulfamethoxazole in a patient with infected cysts caused by a sensitive strain of Escherichia coli. We determined the concentration of trimethoprim and sulfamethoxazole in eight cysts (four proximal, four distal) following therapeutic nephrectomy in a patient treated eight days with trimethoprim-sulfamethoxazole in appropriate doses. In four proximal cysts, mean trimethoprim level was 16.1 +/- 0.8 micrograms/mL with mean sulfamethoxazole level of 94.7 +/- 13.0 micrograms/mL. In distal cysts, mean trimethoprim level was 227.8 +/- 16.8 micrograms/mL with mean sulfamethoxazole level of 9.7 +/- 3.6 micrograms/mL. Serum peak and trough trimethoprim concentrations were 9.8 micrograms/mL and 5.4 micrograms/mL with peak and trough sulfamethoxazole concentrations of 136.0 micrograms/mL and 65.0 micrograms/mL. Significant WBC counts were present in seven cysts, three proximal and four distal. All three proximal cysts were sterile; in contrast, the four distal cysts grew the same strain of E coli isolated from the blood and urine of this patient. The infection resolved following nephrectomy. We conclude that the failure of trimethoprim-sulfamethoxazole to eradicate the infection was caused by the inability of sulfamethoxazole to enter distal cysts in sufficient concentration for the synergistic effect commonly seen with trimethoprim and sulfamethoxazole in combination. Treatment of cyst infections with trimethoprim-sulfamethoxazole should probably be avoided in instances when the organism is resistant to trimethoprim alone.