Sperm-egg plasma membrane fusion during fertilization was studied using guinea pig gametes and mAbs to sperm surface antigens. The mAb, PH-30, strongly inhibited sperm-egg fusion in a concentration-dependent fashion. When zona-free eggs were inseminated with acrosome-reacted sperm preincubated in saturating (140 micrograms/ml) PH-30 mAb, the percent of eggs showing fusion was reduced 75%. The average number of sperm fused per egg was also reduced by 75%. In contrast a control mAb, PH-1, preincubated with sperm at 400 micrograms/ml, caused no inhibition. The PH-30 and PH-1 mAbs apparently recognize the same antigen but bind to two different determinants. Both mAbs immunoprecipitated the same two 125I-labeled polypeptides with Mr 60,000 (60 kD) and Mr 44,000 (44 kD). Boiling a detergent extract of sperm severely reduced the binding of PH-30 but had essentially no effect on the binding of PH-1, indicating that the two mAbs recognize different epitopes. Immunoelectron microscopy revealed that PH-30 mAb binding was restricted to the sperm posterior head surface and was absent from the equatorial region. The PH-30 and PH-1 mAbs did not bind to sperm from the testis, the caput, or the corpus epididymis. PH-30 mAb binding was first detectable on sperm from the proximal cauda epididymis, i.e., sperm at the developmental stage where fertilization competence appears. After purification by mAb affinity chromatography, the PH-30 protein retained antigenic activity, binding both the PH-30 and PH-1 mAbs. The purified protein showed two polypeptide bands of 60 and 44 kD on reducing SDS PAGE. The two polypeptides migrated further (to approximately 49 kD and approximately 33 kD) on nonreducing SDS PAGE, showing that they do not contain interchain disulfide bonds, but probably have intrachain disulfides. 44 kD appears not to be a proteolytic fragment of 60 kD because V8 protease digestion patterns did not reveal related peptide patterns from the 44- and 60-kD bands. In the absence of detergent, the purified protein precipitates, suggesting that either 60 or 44 kD could be an integral membrane polypeptide.