BACKGROUND: Presence of immunocytochemically detectable actin in the rat and mouse sperm head has been enigmatic for years. In this study, we demonstrate actin in the perinuclear theca and show that the detection of actin epitopes in the rat and mouse epididymal spermatozoa can effectively be enhanced by pre-extraction of sperm cells with SDS. METHODS: The study with one monoclonal and one polyclonal anti-actin antibody was carried out at conventional and confocal fluorescence and electron microscope level, and by immunoblotting of proteins isolated from the head and tail fractions. RESULTS: In the head of the control methanol-acetone fixed rat spermatozoa, the polyclonal antibody gave a stronger immunostaining in the postacrosomal area and in the perforatorium than the monoclonal antibody. In the mouse sperm head, the monoclonal antibody labeled the ventral edge of the postacrosomal area and slightly the perforatorium, whereas the polyclonal antibody stained the entire perinuclear space. In the SDS-extracted spermatozoa, an intense postacrosomal and perforatorial labeling was obtained with both antibodies but, in particular in the rat spermatozoa, the middle lateral portion of the postacrosomal segment remained unlabeled. Sonication seemed to cause structural modifications which specifically impeded staining with the monoclonal antibody. Both antibodies detected actin in the basal plate and the monoclonal antibody in the neck. Amorphous matrix of the connecting piece showed immunogold labeling. In the tail, the monoclonal antibody recognized actin and a relatively basic 53 kDa polypeptide, whereas the polyclonal antibody reacted with several protein bands. SDS-soluble actin of the tail was addressed to the midpiece and the SDS-insoluble 53 kDa protein profoundly to the outer dense fibers of the principal piece. CONCLUSIONS: Intense labeling of actin in the SDS-extracted rat and mouse spermatozoa was presumably due to the generated demasking of actin epitopes embedded in the perinuclear cytoplasm. The results are important in confirming that actin in the rat and mouse sperm head is not lost during spermiogenesis but apparently contributes to the three-dimensional packing of the mature perinuclear cytoplasm. This study further demonstrates the importance of the methods used in sample preparation and advantages of confocal microscopy when attempting to detect cytoskeletal proteins which, as in spermatozoa, may occur in small quantities.