Predicting bull fertility is one of the main challenges for the dairy breeding industry and artificial insemination (AI) centers. Semen evaluation performed in the AI center is not fully reliable to determine the level of bull fertility. Spermatozoa are rich in active miRNA. Specific sperm-borne miRNAs can be linked to fertility. The aim of our study is to propose a combined flow cytometric analysis and miRNA profiling of semen bulls with different fertility to identify markers that can be potentially used for the prediction of field fertility. Sperm functions were analyzed in frozen-thawed semen doses (CG: control group) and high-quality sperm (HQS) fraction collected from bulls with different field fertility levels (estimated relative conception rate or ERCR) by using advanced techniques, such as the computer-assisted semen analysis system, flow cytometry, and small RNA-sequencing. Fertility groups differ for total and progressive motility and in the abnormality degree of the chromatin structure (P < 0.05). A backward, stepwise, multiple regression analysis was applied to define a model with high relation between in vivo (e.g., ERCR) and in vitro (i.e., semen quality and DE-miRNA) fertility data. The analysis produced two models that accounted for more than 78% of the variation of ERCR (CG: R2 = 0.88; HQS: R2 = 0.78), identifying a suitable combination of parameters useful to predict bull fertility. The predictive equation on CG samples included eight variables: four kinetic parameters and four DNA integrity indicators. For the HQS fraction, the predictive equation included five variables: three kinetic parameters and two DNA integrity indicators. A significant relationship was observed between real and predicted fertility in CG (R2 = 0.88) and HQS fraction (R2 = 0.82). We identified 15 differentially expressed miRNAs between high- and low-fertility bulls, nine of which are known (miR-2285n, miR-378, miR-423-3p, miR-191, miR-2904, miR-378c, miR-431, miR-486, miR-2478) while the remaining are novel. The multidimensional preference analysis model partially separates bulls according to their fertility, clustering three semen quality variable groups relative to motility, DNA integrity, and viability. A positive association between field fertility, semen quality parameters, and specific miRNAs was revealed. The integrated approach could provide a model for bull selection in AI centers, increasing the reproductive efficiency of livestock.