The use of chromosome analysis on products of conception from spontaneous abortions is recommended to identify a genetic etiology. However, 20% of products of conception cultures are unsuccessful due to microbial contamination or lack of viable dividing cells. Our laboratory implemented a reflex fluorescent in situ hybridization (FISH) assay to detect numeric chromosome abnormalities for unsuccessful cultures. All products of conception samples were simultaneously processed for both chromosome analysis and FISH analysis. If the chromosome analysis was unsuccessful, interphase FISH was performed for chromosomes 13, 16, 18, 21, 22, X, and Y. To assess the performance of the FISH assay, a 3-year retrospective comparative analysis of the FISH results versus chromosome results was performed. Of 5555 total specimens, 4189 (75%) represented chorionic villi/fetal tissue and 1366 (25%) represented tissue of unidentified origin. Of the 1189 tissues of unidentified origin with chromosome or FISH results, 1096 (92%) were XX, indicating that the majority of these tissues are likely maternal in origin. Of the 3361 successful chromosome studies on the chorionic villi/fetal tissue specimens, 1734 (52%) samples had a chromosome abnormality. Of the 762 successful FISH studies on chorionic villi/fetal tissue specimens that were unsuccessful by chromosome studies, 181 (25%) had an abnormal result with the targeted FISH panel. Overall, the FISH panel detected approximately 70% of the chromosome abnormalities in products of conception detectable by karyotype. When the FISH panel results were combined with chromosome analysis for the 4189 chorionic villi/fetal tissue specimens, the overall abnormality rate is 47%. Our reflex FISH assay proved useful for the detection of common chromosome aneuploidies in products of conception samples that failed conventional chromosome analysis. Because of its limited view of the genome, cautious interpretation of FISH results is required for all samples, in particular, trisomy of an acrocentric chromosome, which may represent a Robertsonian translocation. An algorithmic approach to the genetic evaluation of products of conception specimens, with the potential for initial evaluation by a FISH panel, may be warranted.