e17052 Background: Taxane chemotherapies are a mainstay of treatment in mCRPC. The addition of carboplatin (car) to cabazitaxel (cab) has shown efficacy and toxicity within clinical trials. Although car/cab were initiated concurrently in trial, clinical efficacy of car addition in those with PSA progression on cab is unknown. Furthermore, little is known about outcomes in older patients treated with car/cab. Methods: We retrospectively reviewed 72 patients with mCRPC treated with car/cab combination therapy. Demographics, disease characteristics, and treatment history were collected via electronic medical record. Overall survival (OS), progression-free survival (PFS), best overall response (BOR) per RECIST 1.1 in those with measurable disease, PSA30 and PSA50, defined as a decrease in PSA by at least 30% or 50%, respectively, were assessed. Adverse events related to treatment were assessed. In patients ≥ 65 years old, baseline frailty was assessed with the G8 screening tool, which incorporates recent food intake, weight loss, mobility, neuropsychological conditions, BMI, polypharmacy, self-assessed health status, and age. A score of ≤ 14 suggests geriatric vulnerability. Univariate association between clinical variables and the G8 score was assessed by Fisher’s exact test and Chi-square analysis. Survival outcomes were assessed via Kaplan-Meier method. Results: Median age was 66.5 years, 45.8% were Black, and 84.5 % had high volume disease. 88.1% of patients had an impaired G8 score even though 83.7% of all patients included were classified as ECOG 0-1 by their oncologist. Initially, 32 patients (44.4%) started cab monotherapy with car added later and received a median of 4 cycles of cab prior. Median OS was 10.8 months (95% CI 8.8, 13.3) and median PFS was 6.2 months (95% CI 5.8, 8.3). PSA30 and PSA50 were 38.6% and 25.7%, respectively, with BOR of partial response (PR) in 5 patients (8.6%). In those in whom car was added due to a lack of response to cab, 32.3% and 16.1% of patients were able to recapture PSA30 and PSA50 with the combination, respectively; only 1 patient (3.8%) had a PR as BOR. 56.9% of patients did not receive further treatment after car/cab. 93.1% of patients experienced toxicity (90.6% of those in whom car was added) and 34.7% stopped treatment due to toxicity alone. Fatigue (64.8%), anemia (85.9%), leukopenia (36.6%), and thrombocytopenia (40.8%) were the most common. G-CSF support was required in 70.8% and those with an impaired G8 score were significantly more likely to require this support (p=.008). Conclusions: To our knowledge, this is the first study showing that the addition of car in those not responding to cab can recapture PSA response in roughly one third of patients. Since car/cab is toxic, this may represent a strategy to expose only those not responding to cab to the added toxicity of the doublet. More frail patients may experience increased toxicity with car/cab.