Knee osteoarthritis (OA) is a highly prevalent disease and treatment options for early stages of OA are needed. Intraosseous injections of bone substitute and biologic materials have been proposed to expand the therapeutic arsenal by potentially halting OA progression and delaying the need for knee arthroplasty in patients with early/moderate-stage disease. Therefore, the goal of this study was assessed the efficacy and safety of subchondral intraosseous injection for the treatment of knee OA. A systematic review was performed on PubMed-Medline, and the Cochrane Database of systematic reviews. English and Spanish retrospective and prospective studies assessing the results of subchondral intraosseous injection of bone substitute materials and/or biologicals in human patients with knee OA, with a minimum of 6 months of follow-up were collected. A total of 1,081 potential articles were identified through our search. Six studies were included with a total of 163 patients. The mean follow-up was 18 months (range: 6–24 months). Patient reported outcomes measures (PROMs), complications, and conversion to total knee arthroplasty (TKA) were collected. All six studies showed PROMs improvement relative to baseline. Overall, the five studies reporting visual-analog scale (VAS) pain outcomes improved from a baseline mean score of 6.68 to 2.74. Also, knee injury and osteoarthritis score (KOOS), Tegner-Lysholm, and/or international knee documentation committee (IKDC) scores rose compared with baseline scores in all studies. Overall, 2.5% (4/163) of patients had a complication attributed to study-related treatment. Most patients (81%, 86/106) remained TKA-free at a 1-year follow-up. Subchondral intraosseous injections of bone substitute materials and platelet-rich plasma (PRP) suggest (1) improved PROMs of pain and functional status, (2) low complication rate, and (3) relatively low rates of conversion to TKA. However, the current studies investigating these treatments exhibited high degree of heterogeneity in both measurement of outcomes and delivery of treatment, with a high risk of bias. This procedure should not be utilized in advanced knee OA. In light of the limitations of the current literature, advising in favor or against this therapy for early to moderate knee OA is challenging.