There is a trend towards offering immunotherapy to women with unexplained reproductive failure based on abnormal Natural Killer (NK) cell levels. Previous systematic reviews evaluating immunotherapy usage have not focused on women with abnormal level of NK cells. To address the gap in literature, this systematic review aims to evaluate the efficacy of immunotherapy to improve pregnancy outcome in women with recurrent miscarriage (RM) or implantation failure (RIF) specifically selected based on abnormal levels and/or activity of NK cells. Six databases were searched for peer-reviewed studies following PRISMA guidelines. Risk of bias assessment was conducted using RoB2 for randomized controlled trials (RCT) and ROBINS-I for non-RCT. Of 1025 studies identified, seven studies on intravenous immunoglobulin (IVIG) (four), prednisolone (one), etanercept (one) and intralipid (one) were included. Meta-analysis of the non-RCT IVIG studies (557 participants; 312 intervention, 245 controls) showed livebirth in favour of intervention (RR 2.57; 95 % CI = 1.79-3.69; p < 0.05), however there were significant heterogeneity (I2 = 62 %) and moderate to severe risk of bias in these studies. Individual RCTs reported improved livebirth outcome in etanercept, intralipid and prednisolone and this was significant in the former two (p < 0.05). In conclusion, there may be some benefit of immunotherapy, but paucity of high quality evidence means that it is not possible to support the use of immunotherapy even when selected based on abnormal NK cell level/activity. Further research with application of scientifically validated immunological biomarkers in well-planned large scale RCTs will determine whether immunotherapy is beneficial in this subpopulation of women.