Background: Opioid overdose mortality in the United States remains a severe public health crisis, with the burden distributed unequally across communities that differ in epidemic trajectory, baseline resources, and local context. While harm reduction through naloxone distribution and treatment through buprenorphine are both evidence-based strategies, how their effectiveness varies across county-level contexts has received limited quantitative study, limiting the ability of local policymakers to prioritize resources across counties. Methods: We developed a simulation model of opioid use disorder (OUD) progression, calibrated separately to three Pennsylvania counties spanning large urban (Allegheny), mid-sized (Erie), and rural (Clearfield) settings using Bayesian calibration. We projected county-specific overdose mortality trajectories under three levels of proportional increase in dispensing rates of buprenorphine and naloxone (10%, 20%, and 30% above each county's observed baseline dispensing levels), over a 2025-2029 projection horizon. Results: A 30% increase in naloxone dispensing above observed county baseline levels was projected to reduce cumulative overdose deaths over 2025-2029 by approximately 50% in Allegheny County (large urban), modestly in Erie County (mid-sized), and only slightly in Clearfield County (rural). Projected reductions were consistently smaller for buprenorphine across all three counties, except in Erie, where buprenorphine produced larger projected reductions than the other counties. Heterogeneity in naloxone responsiveness was strongly associated with each county's historical naloxone dispensing variability. Conclusions: The same proportional increase in naloxone dispensing yields substantially different projected mortality reductions across counties depending on each county's baseline distribution history, a pattern invisible from mortality statistics alone. County-level context must inform harm reduction and treatment prioritization rather than uniform, population-proportional approaches.