Sclerosing keratitis is the predominant cause of blindness due to onchocerciasis which is a major human parasitic disease caused by the filarial parasite Onchocerca volvulus. In the present investigation, native pathogenic antigens of O. volvulus which are particularly potent in causing interstitial keratitis were characterized utilizing a guinea pig model. Following demonstration of the protein nature of these antigens using pronase digestion, the crude O. volvulus antigen extract was subjected to stepwise procedures of protein purification. At each stage of purification, pooled antigen fractions were injected into one cornea of presensitized guinea pigs followed by clinical evaluation of stromal inflammation and vascularization at different intervals of time after intrastromal challenge. Initial purification of the pathogenic antigens was carried out in the following order: molecular sieve chromatography on Bio-gel A-5m. anion exchange chromatography on Mono Q followed by DEAE-Sepharose CL-6B and cation exchange chromatography on Mono S. Two out of six different pools from the Mono S column (pool a eluted unbound at 10 mM-NaCl and pool e eluted between 130 mM and 475 mM-NaCl) were found to be most pathogenic. Further purification of Mono S pool a and pool e separately by gel filtration chromatography using Superose 12 demonstrated that the fractions which were most potent in inducing interstitial keratitis contained proteins with approximate molecular masses between 100 and 200 kDa. These results show that minor subfractions of total crude antigens of O. volvulus are largely responsible for induction of experimental interstitial keratitis. We have demonstrated the presence of these antigens in O. volvulus microfilariae by their cross-reactivities with anti-microfilarial antibodies, and hence the relevance of the purified antigens to ocular onchocerciasis in man since sclerosing keratitis is associated with invasion of the cornea by O. volvulus microfilariae. Isolation of these two pathogenic antigen pools represents the practical limits of purification and subsequent animal experiments possible with the available amounts of native parasite material obtained from infected human individuals in the absence of a suitable non-human host or of an in vitro culture system for O. volvulus.