It is generally thought that a single ejaculate is a bad predictor of semen quality of a subject, because of significant intra-individual variation. Therefore, we investigated the degree to which the results of a first semen analysis differ from that of a second analysis among men from a general population in Norway. In addition, we analysed how the two different semen results mirrored the overall semen quality assessment. A total of 199 volunteers participated in the study and delivered two semen samples with an interval of 6 months. The semen parameters were determined according to the World Health Organization (WHO) 1999 guidelines, which were also used to determine whether semen quality was normal or abnormal. In addition, the DNA fragmentation index (DFI) was determined using the Sperm Chromatin Structure Assay. The two samples from each individual were very similar with regard to standard semen parameters and DFI (r(s:) 0.67-0.72), and there were no significant systematic differences between the two samples. The result of the first sample (normal/abnormal) was highly predictive of the overall conclusion based on the two samples (sperm concentration: in 93% of the cases (95% confidence interval [CI]: 89%-96%); sperm motility: in 85% of the cases (95% CI: 79%-89%); overall semen quality: in 85% of the cases (95% CI: 80%-90%). In epidemiological studies, one ejaculate is a sufficient indicator of semen quality in a group of subjects. In a clinical situation, when the question is whether the semen quality is normal or not, the first ejaculate will, in at least 85% of cases, give a correct overall conclusion.