We tested the ability of chromosomes in a mitotic cytoplasm to organize a bipolar spindle in the absence of centrosomes. Sea urchin eggs were treated with 5 X 10(-6) colcemid for 7-9 min before fertilization to block future microtubule assembly. Fertilization events were normal except that a sperm aster was not formed and the pronuclei remained up to 70 microns apart. After nuclear envelope breakdown, individual eggs were irradiated with 366-nm light to inactivate photochemically the colcemid. A functional haploid bipolar spindle was immediately assembled in association with the male chromosomes. In contrast to the male pronucleus, the female pronucleus in most of these eggs remained as a small nonbirefringent hyaline area throughout mitosis. High- voltage electron microscopy of serial semithick sections from individual eggs, previously followed in vivo, revealed that the female chromosomes were randomly distributed within the remnants of the nuclear envelope. No microtubules were found in these pronuclear areas even though the chromosomes were well-condensed and had prominent kinetochores with well-developed coronas. In the remaining eggs, a weakly birefringent monaster was assembled in the female pronuclear area. These observations demonstrate that chromosomes in a mitotic cytoplasm cannot organize a bipolar spindle in the absence of a spindle pole or even in the presence of a monaster. In fact, chromosomes do not even assemble kinetochore microtubules in the absence of a spindle pole, and kinetochore microtubules form only on kinetochores facing the pole when a monaster is present. This study also provides direct experimental proof for the longstanding paradigm that the sperm provides the centrosomes used in the development of the sea urchin zygote.