BACKGROUND: Cancer caregivers are at high risk for distress, yet underutilize supportive care due to access and time barriers. We developed and psychometrically evaluated CancerSupportSource®-Caregiver (CSS-CG), part of a web-based distress screening and referral program. The goal of CSS-CG is to identify and respond to unmet caregiver need, enabling caregivers to better navigate care, practice self-care, and support patients. METHOD: 400 caregivers enrolled in Cancer Support Community’s Cancer Experience Registry and rated 33 CSS-CG items (0=Not at all; 4=Very seriously) and validation measures (PROMIS-29 v 2.0; PHQ-4; NCCN DT, ZBI-12, CRA). Scale reduction was informed by item endorsement, iterative exploratory / confirmatory factor analysis (EFA; CFA), Pearson correlations, and clinician input. Accuracy of 2-item depression and anxiety risk scales were assessed with AUC and sensitivity/specificity analysis. RESULTS: Participants were 81% women; Mean Age=51.1 (range=20-89; 16% age 65+); 83% Non-Hispanic White, 6% Hispanic; 64% current caregivers. 42% cared for spouses; 21% parents. Years of care provided: 20% <1y, 28% 1-2y, 17% 3-5y, 9% 6-10y, 6% 11+y. 400 people completed ≥32/33 distress items and were randomly split into two samples: EFA(N=250); CFA(N=150). After removing items with limited variability, low endorsement, low factor loadings (<.30), or low validation measure correlations, 18 items remained across 5 factors plus 1 substance use item: Emotional Well-being (5 items, α=.89), Caregiving Tasks (4 items, α=.90), Finances (2 items, α=.88), Healthy Lifestyle (3 items, α=.81), and Patient Well-Being (4 items, α=.87). EFA explained 67% of model variance, RMSR=0.02, TLI=0.95, RMSEA=0.06; CFA: CFI=0.94, TLI=0.93, RMSEA=0.07. CSS-CG had high internal reliability (α=.92) and strong convergent validity: greater total distress was associated with poorer QOL (rs=-.23-.66), greater NCCN DT distress (r=.61), and greater caregiver burden (r=.57). The 2-item depression and anxiety risk scales demonstrated high sensitivity (.87-.95) and adequate specificity (.63-.68). CONCLUSION: CSS-CG is a psychometrically robust measure of caregiver distress. In its full program implementation, it can facilitate distress screening, referral, and follow-up to rapidly assess caregivers’ unmet needs and enhance well-being. Future implementation goals include understanding how to best engage caregivers in diverse settings, including oncology practices and ambulatory sites.