There is ongoing controversy about the recommended dose of cabazitaxel in patients with metastatic castration-resistant prostate cancer (mCRPC). This multicenter phase II open-label, randomized, parallel-group study compared 3-weekly cabazitaxel at 25 mg/m2 (conventional arm A) with cabazitaxel therapeutic drug monitoring (experimental arm B) in mCRPC. The primary objective was to improve the clinical feasibility rate (CFR), defined as the absence of grade 4 neutropenia or thrombocytopenia, any thrombocytopenia with bleeding, febrile neutropenia, severe nonhematologic toxicity, withdrawal for cabazitaxel-related toxicity, or death. A total of 60 patients had to be randomized to detect a difference in CFR of 35% (power 80%, two-sided alpha 10%). A total of 40 patients were randomized to arm A and 33 patients to arm B. CFR was 69.4% in arm A and 64.3% in arm B (P = 0.79). Week-12 PSA response was 38.5% in both arms. A radiological response by RECIST v.1.1 was seen in 3 (9.7%) patients in arm A versus 6 (23.1%) patients in arm B (P = 0.28), disease progression was higher in arm A compared with arm B (61.3% vs. 30.8%, P = 0.05). Median progression-free survival was longer in arm B compared with arm A (9.5 vs. 4.4 months; HR = 0.46; P = 0.005). Median overall survival was higher in arm B compared with arm A (16.2 vs. 7.3 months; HR = 0.33; P < 0.0001). Pharmacokinetic-guided dosing of cabazitaxel in patients with mCRPC is feasible and improves clinical outcome due to individual dose escalations in 55% of patients.