Some recent studies have indicated that sperm concentration has decreased during the last 50 years. However, comparisons between laboratories have revealed that geographical differences seem to exist and that any decrease may not be global. One point of criticism concerning comparison of results from different laboratories has been that some of the discrepancies detected could reflect the lack of standardized methods used in the different laboratories. Four teams, each consisting of one physician and one technician from groups which have recently published data on semen quality, met in order to evaluate the variability between their laboratories on semen analysis. Twenty-six fresh semen samples from unselected men were analysed. The groups analysed the samples according to the normal practice in their laboratories, using their own equipment. The variation between laboratories was estimated through a random effects model. For sperm concentration and semen volume assessment a remarkable consistency between laboratories was detected, in contrast to the very considerable inter-individual variation. For sperm motility and morphology assessments interlaboratory consistency was much poorer. In conclusion, evaluation of sperm motility and morphology characteristics requires further standardization in order to achieve comparable data from different laboratories. However, semen volume and sperm concentration are characteristics which can be compared reliably between laboratories, when similar methodologies are used.