In a prospective randomized double-blind trial, pivmecillinam was compared with cotrimoxazole (TMP-SMX), both given orally for a period of 5 days, for the treatment of 59 children with shigellosis. 29 patients were treated with pivmecillinam and 30 with cotrimoxazole. 14% of shigella organisms isolated were resistant to pivmecillinam and 21% to TMP-SMX. The diarrhea persisted for a mean (+/- SD) period of 74 +/- 24.8 h in the pivmecillinam-treated patients versus 73.8 +/- 34 h in the TMP-SMX-treated patients. Duration of fever, positive stool culture, visible blood, occult blood, and pus cells in the stools were similar for both treatment groups. Five patients (17%) in the pivmecillinam group and 4 patients (13%) in the cotrimoxazole group fulfilled the clinical criteria that defined treatment failure. One patient (3.4%) in the pivmecillinam group and 2 (6.6%) in the TMP-SMX group evidenced recurrence of the diarrheal symptoms at the follow-up visit. No major drug-related side effects were observed in either group. We concluded that pivmecillinam is equivalent to cotrimoxazole in the treatment of shigellosis in children.