To differentiate round cells in semen samples of subfertile men and evaluate the clinical significance during infertility investigation. One hundred and eight randomly chosen couples with a median duration of infertility of 4 (range, 1 to 20) years presenting at the outpatient infertility clinic of the University of Heidelberg, Germany. Differentiation of round cells in semen by means of monoclonal antibodies (mABs) and a streptavidin-biotin system for staining. Correlation of results with medical history, outcome of clinical examination, sperm analysis, microbial screening of both partners, evaluation of sperm functional capacity in vivo by means of the postcoital test (PCT) and in vitro with the standardized crossed sperm-cervical mucus penetration test (SCMPT) and the subsequent fertility in a prospective study. The method used for differentiation of round cells proved to be practical and suitable for routine use. The percentage of leukocytes ranged from 0% to 58% with a median of 3%. Number of round cells and percentage of leukocytes did not differ markedly with regard to andrologic history, clinical findings, for example, varicocele, results of standard sperm analysis, and microbial colonization of semen samples. However, high rates of leukocytes of the round cells correlated with reduced sperm count and morphology and results of PCT. Leukocyte-positive (> 15% leukocytes) specimens were also significantly more frequent in case of inadequate SCMPT and reduced sperm penetration ability in vitro. In asymptomatic patients (in terms of genital tract infection), the majority of round cells consist of immature germ cells and < 5% are white blood cells. The streptavidin-biotin system and the mABs used in this study proved to be useful to identify patients with elevated rates of leukocytes in semen possibly reflecting subclinical genital tract infection with influence on sperm functional capacity and subsequent fertility. Thus the procedure can be recommended to be included in a comprehensive evaluation of male fertility.