Abstract Study question Can paternal risk factors be minimized and prevented in ICSI procedure? Summary answer The influence of DSBs can be prevented in ICSI cases with normal ovarian response, whereas DSB is exclusively associated with miscarriage in poor ovarian reserve. What is known already Sperm DNA fragmentation (SDF) is intricately linked to male fertility capacity, with a high prevalence affecting 20% of men with unexplained infertility and up to 50% of those with idiopathic infertility. Despite these statistics, the association between paternal DNA damages and clinical outcomes is controversial. Variations in SDF effects may be influenced by female factors associated with maternal DNA repair activity. To elucidate this interrelationship, we assessed SDF effects on diverse ovarian responses using POSEIDON stratification. The goal was to identify potential paternal risk factors for ART as a preventative solution and devising preventative strategies. Study design, size, duration This retrospective study included 323 couples who underwent ICSI cycles at Lee Women’s Hospital. Normal responders (n = 89) were identified based on female age (<35 years) and AMH levels (≥1.2 ng/mL). Patients (n = 234) fulfilling POSEIDON group 1–4 criteria were categorized as poor ovarian response (POR). Clinical outcomes assessed included clinical pregnancy, miscarriage, and live birth rates. Participants/materials, setting, methods Paternal risk factors include male age, semen parameters (CEROS II CASA, Hamilton-Thorne), and two types of SDF evaluation. LensHooke® R10 kit (Bonraybio, Taiwan) was used to detect total SDF (tSDF), which includes single-strand breaks and double-strand breaks (DSBs). LensHooke® R11 kit was specifically used for DSB detection. Kruskal-Wallis test was conducted to compare controls and cases of ART failure. Univariate and multivariate logistic regression analyses were employed to identify potential factors. Main results and the role of chance We sought to evaluate paternal effects during the early gestation stage by comparing parental characteristics in groups with clinical pregnancy and those without pregnancy. No significant differences in paternal indications were observed between the groups. Notably, significantly lower AMH levels and advanced maternal age were prevalent in pregnancy failures across both normal responders and POR cases, highlighting maternal factors’ dominance in these outcomes. When dissecting paternal effects during the later gestation stage, male factors did not show a significant difference between the group with live birth and the group with miscarriage among normal responders. Nevertheless, several paternal pathogenic indications were identified between the groups in POR. Miscarriages were associated with compromised sperm total motility (68% vs. 54.5%, p = 0.01), reduced normal morphology rate (5.5% vs. 2.5%, p = 0.03), and elevated DSB levels (7.5% vs. 18.8%, p = 0.01). Furthermore, multivariate regression analysis identified high DSBs as an exclusive paternal factor in miscarriage risk (Odds ratio: 1.19, 95% confidence interval 1.04–1.36, p = 0.01). In the ROC curve, a DSBs threshold of 19%, based on the R11 technique, offers significant value in predicting miscarriage with an accuracy of 81%. Limitations, reasons for caution The limitations of the current study are single-center trial, and relatively small sample size. Further clinical trials are needed to robustly establish the relationship between DSB and miscarriage. Wider implications of the findings The impact of DSBs is overlooked in normal responders but not in POR due to different levels of maternal DNA repair activity. These results advocates for (1) the necessity of DSBs (R11) assessment for POR, and (2) egg donation as a potential strategy for couples facing high DSBs and POR. Trial registration number CS2-20012