Background: Somatostatin receptors (SSR) are overexpressed by neuroendocrine tumors (NETs). Peptide receptor radionuclide therapy (PRRT) targets the tumor cells through SSRs. Prediction of efficacy is critical since SSR expression or chromogranin A (CgA) levels have limited (20-60%) predictive ability. We assessed the utility of a blood-based multigene NET transcript analysis (NETest) to predict PRRT efficacy in 2 PRRT-treated NET cohorts. The algorithm comprises > 50 genes + clusters defining growth factor signaling (GFS) and metabolism (MTB). Methods: We investigated 2 independent ¹⁷⁷Lu-PRRT-treated groups: #1 (n= 72), median total activity 18.5GBq (6.5-27.8) and #2 (n= 24), 24.5GBq (11.4-29.7). Baseline evaluations: grade (Ki67), SSR imaging (SRI), CgA (ELISA), and NETest (qRT-PCR - multianalyte algorithmic analyses). Using #1 we mathematically devised a predictive response index (PRI) of > 50 NET genes, MTB and GFS clusters & Ki67). We then assessed this PRI for PRRT efficacy in #2. RECIST criteria were used to assess disease control (responder vs non-responder). Statistical analyses: multiple regression, Kaplan-Meier survival, Chi². Results: Cohorts were comparable by grade, staging (IV) and SRI uptake. #2 had more advanced disease (progressive 100% vs. 74%, p< 0.01). At restaging, disease control rates were similar (68% and 83%, p= 0.23) and median PFS was not reached in either cohort (follow-up 15 and 9.5 months, respectively). Baseline CgA was not different (644±133 vs. 684±295ng/ml). NETest levels were significantly elevated in #2 (67±6% vs. 51±3%, p= 0.03), consistent with the progressive disease profile. PRRT response was not predicted by grading (p= 0.15), SRI (p= 0.41) or CgA (p= 0.12). In #1 the PRI accurately (> 90%) predicted responders (95%) and non-responders (90%), significantly better than SRI (Level 4: 43% accuracy: Chi²= 41, p< 0.0001). In #2 the PRI was also predictive (83%) and significantly more accurate than SRI (46%, Chi²= 21, p< 0.01). Conclusions: NET gene analysis in blood significantly outperformed (p< 0.01) CgA, Ki67 and SRI for prediction of PRRT efficacy. PRRT efficacy can be accurately predicted (83-94%) by pre therapy blood transcript analysis.