Although infection following intradiscal injections has been recognized as a distinct entity, discitis following chemonucleolysis has been often attributed to a chemical reaction from chymopapain. In the first part of this study the effect of chymopapain and Conray 280 on a wide range of bacteria was measured in vitro. Chymopapain was found to have a bactericidal effect on all bacteria tested, which was more pronounced with gram positive organisms, whereas Conray 280 showed very little if any antibacterial effect after 48 hours. The aim of the second part of the study was to test the hypothesis that discitis following intradiscal chymopapain injection is due to infection and not to a chemical reaction. Multiple level lumbar intradiscal injections were carried out in eight mature sheep. Sixteen discs in four sheep were injected with a mixture of reconstituted chymopapain and a Staphylococcus epidermidis suspension. Sixteen discs in another four sheep were injected with reconstituted chymopapain only. All sheep were sacrificed at 6 weeks and the discs and end-plates were examined radiologically, and by histopathology and nuclear material was cultured for bacteria. None of the controls showed any evidence of discitis, whereas all sheep injected with bacteria had typical radiologic and histopathologic changes of discitis. However, in most cases in which end-plate lesions were well established there was no evidence of bacteria at sacrifice. These findings support the opinion that discitis following intradiscal injection is always due to infection introduced by the needle tip.