Objectives To investigate which infertile men with semen parameters above WHO reference limits at first semen analysis deserve a second semen test. Materials and methods Data from 1358 consecutive infertile men were analysed. Patients underwent two consecutive semen analyses at the same laboratory. Descriptive statistics and logistic regression models tested the association between clinical variables and semen parameters. A new predicting model was identified through logistic regression analysis exploring potential predictors of semen parameters below WHO reference limits after a previously normal one. Diagnostic accuracy of the new model was compared with AUA/ASRM and EAU guidelines. Decision curve analyses (DCA) tested their clinical benefit. Results Of 1358, 212 (15.6%) infertile men had semen parameters above WHO reference limits at first analysis. Of 212, 87 (41.0%) had a second semen analysis with results above WHO reference limits. Men with sperm parameters below reference limits at second analysis had higher FSH values, but lower testicular volume (TV) (all p<0.01) compared to men with a second semen analysis above WHO limits. At multivariable logistic regression analysis, lower TV (OR 0.9, p = 0.03), higher FSH (OR 1.2, p<0.01), and lower total sperm count (OR 0.9, p<0.01) were associated with second semen analyses below WHO limits. DCA showed the superior net benefit of using the new model, compared to both AUA/ASRM and EAU guidelines to identify those men with a second semen sample below WHO limits after a previously normal one. Conclusions Approximately 60% of infertile men with a first semen analysis above WHO limits have a second analysis with results below limits. The newly identified risk model might be useful to select infertile men with initial semen results above WHO limits who deserve a second semen analysis.