The results of a controlled therapeutic trial comparing 2 groups of patients presenting with hemolytic-uremic syndrome (HUS) are reported. Group A (15 children) was given urokinase (UK) and heparin; group B (18 children) received no treatment. Ages of patients, the delay before admission, the severity of anemia, thrombocytopenia and initial renal failure were similar in both groups. UK was responsible for bleedings in 12 children, minimal in 8, severe in 4. No child died in group A, 3 children died in group B (n.s.). Durations of hemolysis, thrombocytopenia and anuria were similar in both groups. Long-term evolutions of renal function and arterial pressure were comparable in both groups. Needle kidney biopsy (26 cases) showed cortical necrosis in 3 children of group A and in 2 of group B, and glomerular thrombotic microangiopathy in 10 children of group A and in 11 of group B. The average ratio of injured glomeruli was 40 (19 to 80) in group A, and 38 (21 to 75) in group B. Two children in group A and 3 children in group B presented with 50 to 80% of glomerular lesions. This trial suggests that UK is of no significant value in the treatment of HUS.