The forms and locations of centrosomes in mouse oocytes and in sea urchin eggs were followed through the whole course of fertilization and first cleavage by immunofluorescence microscopy. Centrosomes were identified with an autoimmune antiserum to centrosomal material. Staining of the same preparations with tubulin antibody and with the DNA dye Hoechst 33258 allowed the correlation of the forms of the centrosomes with the microtubule structures that they generate and with the stages of meiosis, syngamy, and mitosis. The results with sea urchin eggs conform to Boveri's view on the paternal origin of the functional centrosomes. Centrosomes are seen in spermatozoa and enter the egg at fertilization. Initially, the centrosomes are compact, but as the eggs enter the mitotic cycle the forms of the centrosomes go through a cycle in which they spread during interphase, apparently divide, and condense into two compact poles by metaphase. In anaphase, they spread to form flat poles. In telophase and during reconstitution of the daughter nuclei, the centrosomal material is disposed as hemispherical caps around the poleward surfaces of the nuclei. Mouse sperm lack centrosomal antigen. In the unfertilized mouse oocyte, the meiotic spindle poles are displayed as broad-beaded centrosomes. In addition, centrosomal material is detected in the cytoplasm as particles, about 16 in number, which are foci of small aster-like arrays of microtubules. The length and number of astral microtubules correlate with the size of the centrosomal foci. After sperm incorporation, as the pronuclei develop and more cytoplasmic microtubules assemble, a few of the foci associate with the peripheries of the nuclei. The number of foci multiplies during the first cell cycle. At the end of interphase, all of the centrosomal foci have concentrated on the nuclear peripheries and the cytoplasmic microtubules have disappeared. At prophase, the centrosomes are seen as two irregular clusters, marking the poles which, at metaphase and anaphase, appear as rough bands with foci, and the spindle is typically barrel-shaped. At telophase, the centrosomes are seen as arcs that lie on the nuclear peripheries after cleavage. The ordering of microtubules in all the stages reflects the shapes of the centrosomes. The findings on the sea urchin confirm the classical theory of the paternal origin of centrosomes and contrast with observations tracing the mitotic poles of the mouse egg to maternal centrosomal material. This evidence strengthens the conclusion that mouse centrosomes derive from the oocyte.