Necrotizing soft tissue infections are classified by the type of infecting organism, the presenting clinical picture, and the treatment required. However, reliance on this schema is impractical since it often leads to an inordinate delay in appropriate surgical therapy with an unwarranted loss of a limb or life. Since 1958, 21 patients were treated at the UCLA Medical Center with necrotizing soft tissue infections. Unifocal ulcerations and nonspreading infections were excluded. Of the 21 patients, the initial classification of the infections included necrotizing fasciitis in 38 percent, clostridial gangrene in 38 percent, bacterial synergistic gangrene in 19 percent, and streptococcal hemolytic gangrene in 5 percent. Diabetes or evidence of immunosuppression was found in 71 percent of the patients. The course could be traced to either a perforated viscus in 43 percent or a traumatic injury in 43 percent. No single clinical sign, including tissue gas, was diagnostic for a specific type of necrotizing soft tissue infection. Culture revealed a polymicrobial flora in 76 percent. Overall mortality was 52 percent and the amputation rate was 36 percent. Mean time to appropriate surgical therapy was 1.9 days. Operations performed more than 24 hours after recognition of infection resulted in a 70 percent mortality versus a 36 percent mortality when operations were performed less than 24 hours after recognition. A lesser operation to conserve tissue resulted in a 71 percent mortality versus a 43 percent mortality with initial radical surgery which encompassed all devitalized tissue. Based on these data, we conclude that classification of necrotizing soft tissue infections should be simplified to clostridial and nonclostridial infections. Radical operative debridement, regardless of tissue loss, should be carried out immediately after fluid resuscitation, and antibiotic coverage must be broad spectrum from the time of onset due to the polymicrobial nature of these infections.