The distribution of alpha- and gamma-tubulin in human and mouse oocytes has been investigated immunocytochemically. Comparisons have been made between freshly recovered and aged oocytes (both human and mouse), and also between human oocytes before and after exposure to cryoprotectant. Control fresh human oocytes had compact anastral spindles oriented orthogonal to the oolemma, with the pole adjacent to the oolemma being smaller than that directed towards the centre of the oocyte. Each pole was associated with a ring of particulate gamma-tubulin staining that extended a short distance into the body of the spindle. No alpha- and gamma-tubulin staining was found elsewhere in the ooplasm. Human oocytes which had failed to fertilize after an 18 h incubation with spermatozoa and had spent a further 6-8 h in culture showed an increased incidence of spindle abnormalities and of the proliferation of ooplasmic microtubules, which became more pronounced with age post-ovulation. The gamma-tubulin staining pattern of these aged human oocytes revealed greater staining over the whole of the spindle than in fresh oocytes. Examination of mouse oocytes aged in vitro or in vivo showed similar evidence of microtubule proliferation and disorganization, and the gamma-tubulin staining pattern was a sensitive indicator of ageing. The spindles of most fresh human oocytes exposed to 1.5 M dimethyl sulphoxide (DMSO) at 4 degrees C differed from controls in being slightly reduced in size or in having more pointed spindle poles with smaller diameters, both indications that some dismantling of the microtubules had occurred. The distribution of gamma-tubulin in these oocytes extended over more of the spindle. Restoration of DMSO-exposed oocytes to control medium at 37 degrees C for an extended period restored spindle structure to a state closely resembling that in controls. However, recovery of an exclusively polar gamma-tubulin staining did not occur. In both controls and DMSO-exposed human oocytes, chromosomes were arranged on the metaphase equatorial plate. In contrast, exposure of oocytes to 4 degrees C in the absence of DMSO caused dismantling of the spindle. It is concluded that (i) changes in microtubule organization with ageing of oocytes makes them unsuitable for use therapeutically after re-insemination or intracytoplasmic sperm injection, (ii) conditions of cryoprotectant addition previously found optimal for the stabilization of the spindle in the mouse oocyte also appear to be effective in stabilizing the spindle of the human oocyte, and (iii) the distribution of gamma-tubulin in relation to the spindle of the human oocyte appears to be sensitive to age and conditions.