Background and purpose — Platelet-rich plasma (PRP) is broadly used in the treatment of knee osteoarthritis, but clinical outcomes are highly variable. We evaluated the effectiveness of intra-articular injections with Autologous Conditioned Plasma (ACP), a commercially available form of platelet-rich plasma, in a tertiary referral center. Second, we aimed to identify which patient factors are associated with clinical outcome. Patients and methods — 140 patients (158 knees) with knee osteoarthritis (Kellgren and Lawrence grade 0–4) were treated with 3 intra-articular injections of ACP. The Knee Injury and Osteoarthritis Outcome Score (KOOS), pain (Numeric Rating Scale; NRS), and general health (EuroQol 5 Dimensions; EQ5D) were assessed at baseline and 3, 6, and 12 months’ follow-up. The effect of sex, age, BMI, Kellgren and Lawrence grade, history of knee trauma, and baseline KOOS on clinical outcome at 6 and 12 months was determined using linear regression. Results — Mean KOOS increased from 37 at baseline to 44 at 3 months, 45 at 6 months, and 43 at 12 months’ follow-up. Mean NRS-pain decreased from 6.2 at baseline to 5.3 at 3 months, 5.2 at 6 months, and 5.3 at 12 months. EQ5D did not change significantly. There were no predictors of clinical outcome. Interpretation — ACP does not lead to a clinically relevant improvement (exceeding the minimal clinically important difference) in patients suffering from knee osteoarthritis. None of the investigated factors predicts clinical outcome.