Cytogenetic studies on spontaneous abortions traditionally have used one of two methodologies, direct preparations or long-term culture, to determine the chromosome constitution of either the cytotrophoblast or villous stroma, respectively. Few studies have utilized both techniques simultaneously to compare the relative efficiencies of each method and to assess the contribution of confined placental mosaicism (CPM). The present report summarizes cytogenetic studies on 691 consecutive spontaneous abortions using long-term culture, direct preparations, or both. All 691 cases were analyzed by long-term culture and 177 cases were analyzed using both long-term culture and direct preparations. The results indicate that the two methods have similar success rates, 82% for long-term culture and 76% for direct preparation; however, the proportion of normal females was significantly increased in the culture method, presumably attributable to maternal contamination. In 107 cases, results were obtained from both methods with 22 discrepancies identified. However, most of these involved a 46,XX result in culture, consistent with maternal contamination in the cultured preparation. Therefore, to estimate the proportion of CPM we excluded cases with a 46,XX result in culture and found four (6.1%) of the remaining 65 cases to be consistent with CPM. These cases consisted of normal or mosaic aneuploid cytotrophoblast and non-mosaic aneuploid villous stroma. These studies suggest that each method has specific advantages in the analysis of spontaneous abortions. Direct preparations are less prone to maternal contamination, but certain chromosome abnormalities are more likely to be identified using long-term culture.