The limitations of semen analysis as routinely practiced by light microscopy are briefly reviewed. The value of using SEM together with information obtained from TEM has been examined in a ‘blind’ study on three men attending a sub-fertility clinic so that clinical details, including sperm concentration, were not known to the investigators until the work was completed; the men had a clinical history of infertility and the sperm concentrations of the samples examined by electron microscopy were 16 × 10 6/ml, 84 × 10 6/ml and 237 × 10 6/ml. For study by SEM the cells were fixed in buffered glutaraldehyde, concentrated by low-speed centrifugation and tried with acetone, collected on a Millipore filter, dried by the critical point method and coated with gold. This method of preparation gave excellent preservation. Micrographs were obtained showing up to 40 spermatozoa per field which were suitable for assessing variations in the general shape and size of the cells and, at high magnification, differences in surface texture. All the samples contained spermatozoa possessing a large mass of extraneous cytoplasm which was readily visible by SEM. When examined by TEM this region possessed features characteristic of the residual cytoplasm (or body) of spermatids. This results, together with others reported in the present paper and data published elsewhere, suggests that sperm dysgenesis may be a relatively common feature of the spermatozoa of sub-fertile men. It is suggested that the degree of differentiation could be used as an objective basis for identifying and hence classifying, cells considered to be abnormal. It is concluded that SEM, together with information obtained by TEM, can make a valuable contribution to semen analysis of sub-fertile men with widely different sperm concentrations.