Thirty-two horses with clinicopathologic evidence of endotoxic shock were randomly selected for a double-blind trial of hyperimmune lipopolysaccharide (LPS) core antigen plasma. Horses were suffering from acute toxic enteritis (n = 15), 360 degrees volvulus of the large colon (n = 9), proximal jejunitis/duodenitis (n = 6), or strangulating obstruction of the small intestine (n = 2). Plasma was harvested from suitable equine plasma donors (preimmune plasma) and horses were immunized with a whole-cell bacterin of an Rc mutant E. coli (J5). Plasma was again harvested from these horses when IgG ELISA titers recognizing LPS core antigen were greater than 1:32,000. All horses included in the trial received either preimmune or hyperimmune plasma in addition to traditional therapy (fluids, antimicrobials, antiinflammatory agents, etc.) as dictated by the attending clinician. The mortality rate in the group of horses receiving hyperimmune plasma was 13%; in the control group receiving preimmune plasma, the mortality rate was 47% (P = .045). Horses receiving J5 hyperimmune plasma had a significantly improved clinical appearance 48 hours after plasma administration (P less than .05) and a shorter period to recovery than control horses (P = .069).