The development and application of fluorescence in-situ hybridization (FISH) has opened the way for comprehensive studies on numerical chromosome abnormalities in human spermatozoa. FISH can be rapidly applied to large numbers of spermatozoa and thus overcomes the major limitation of karyotyping spermatozoa after penetration of zona-free hamster oocytes. The simultaneous hybridization of two or more chromosome-specific probes to spermatozoa and subsequent detection of the bound probes using different fluorescent detection systems enables two or more chromosomes to be localized simultaneously in the same spermatozoon and provides a technique for undertaking reasonable estimates of aneuploidy. The most commonly used probes are those which bind to the centromeric region of specific chromosomes. Most studies to date have concentrated on estimating aneuploidy in spermatozoa from normospermic men, although reports are beginning to appear on aneuploidy in spermatozoa from subfertile and infertile men. Multi-probe FISH studies have generally reported disomy (hyperhaploidy) estimates of 0.05-0.2% per chromosome. There is preliminary evidence that some chromosomes such as X, Y and 21 are predisposed towards higher rates of non-disjunction during spermatogenesis. There are also suggestions of inter-donor variability in aneuploidy frequencies for specific chromosomes, although this requires confirmation in larger studies. While FISH is clearly a powerful technique that has many applications in reproductive medicine, it must also be realized that it does have limitations and the technology itself is still evolving and has yet to be fully validated on spermatozoa.