40 Background: LuPSMA is a newly established treatment in patients with mCRPC, but PSA and survival outcomes vary widely, and predictors of treatment responses are needed. LuPSMA delivers radiation to tumor tissues that can also be imaged with SPECT/CT which provides semiquantitative estimates of TTV and identification of NLs. This study investigates the use of SPECT/CT in early cycles to predict LuPSMA outcomes. Methods: Between June and December 2022, mCRPC patients who initiated the 2nd cycle of LuPSMA with SPECT/CT 24 hours post-treatment were retrospectively reviewed. We evaluated associations between TTV and the appearance of NLs at the start of the 2nd or 3rd cycle with PFS and OS using Cox regression analysis. NLs were classified as non-PSMA-avid with no or low uptake (< liver) with corresponding findings on CT and PSMA-avid (uptake > liver). All analyses were adjusted for change in PSA relative to baseline. Results: Sixty-six mCRPC patients (median age, 73.5) received a median of 4 (IQR: 3-5) cycles of LuPSMA. Median follow-up starting at 2nd cycle was 26 weeks (IQR: 21-36) with 47/66 (71%) alive at the time of the analysis. A reduction of ≥50% in PSA (PSA50) was noted in 33/66 (50%) patients. Changes in PSA and TTV at 2nd and 3rd cycle were apparently concordant in 50/66 (76%) and 42/51 (82%) and were significantly correlated (r= 0.55 and 0.56, both p<0.001). Patients with higher absolute TTV adjusted for PSA change at 2nd cycle had worse PFS and OS (HR=1.3 and 2.26) with consistent results at 3rd cycle (Table). NLs were detected in 13/66 (7/13 PSMA-avid, 6/13 non-PSMA-avid) and 7/51 patients (5/7 PSMA-avid, 2/7 non-PSMA-avid) at 2nd and 3rd cycle, respectively. Patients with NLs at 2nd cycle had higher risks of progression and death (HR=4.53 and 6.28). Conclusions: Higher SPECT/CT TTV at 2nd and 3rd cycle and detection of NLs at 2nd cycle were associated with higher risks of progression and death. Although changes in PSA and TTV were correlated, absolute SPECT/CT TTV in early cycles provided a complementary ability to predict outcomes of LuPSMA. [Table: see text]