Preselection of the gender of offspring is a subject that has held man's attention since the beginning of recorded history. Most scientific hypotheses for producing the desired sex of offspring address separation of X- and Y-bearing sperm, and most have had limited, if any success. Eight of these hypotheses and their experimental verifications are discussed here. Three hypotheses are based on physical characteristics of sperm, one on supposed differences in size and shape, another on differences in density, and a third on differences in surface charge. There has been no experimental verification of differences based on size and shape, and the results from attempts to verify separation of X- and Y-bearing sperm based on density have been mixed. Electrophoresis may provide a method for separating X- and Y-bearing sperm, but it is currently unproven and would be of little practical utility, since sperm motility is lost. A fourth hypothesis employs H-Y antigen to select preimplantation embryos. This method reliably produces female offspring, but does not permit the selection of male offspring and does not work on sperm. There are two applications of the theory that X- and Y-bearing sperm should be separable by flow fractionation. Flow fractionation using thermal convection, counter-streaming sedimentation, and galvanization is highly promoted by its originator but has not gained wide acceptance due to lack of independent confirmation. Flow fractionation by laminar flow is said to provide up to 80% enrichment of both X- and Y-bearing sperm; however, this method also has not been confirmed by other workers or tested in breeding trials. The sixth theory discussed is that of separation through Sephadex gel filtration. This method may provide enrichment of X-bearing sperm, but, again, other experimenters have not been able to adequately confirm the enrichment. The best-known approach to sperm separation is that employing albumin centrifugation, yet even with this method, not all researchers have been able to confirm a final fraction rich in Y sperm, and trials in animals have given contradictory results. The most reliable method for separating X- and Y-bearing sperm is use of flow cytometric and flow sorting techniques. These techniques routinely separate fractions with a purity greater than 80% and can be above 90%. Unfortunately, these methods do not always work for human samples. Furthermore, as with electrophoretic approaches, the methods identify and separate only chemically fixed sperm and provide limited biological applications.(ABSTRACT TRUNCATED AT 400 WORDS)