Opioid overdose has claimed the lives of over 340,000 Americans in the last decade. Over that same period, policymakers have taken steps to increase the availability of naloxone-an opioid antagonist used to rescue overdose victims-to people in the community. Previous studies, most of which have examined the effects of state laws designed to facilitate access to naloxone, have reached mixed conclusions about the effects of naloxone access on fatal and non-fatal overdoses. This paper exploits a unique policy experiment provided by two naloxone giveaways intended to increase naloxone possession among the general public in Pennsylvania to estimate the causal impact of naloxone distribution on fatal overdoses and opioid-related emergency department (ED) visits. Using a difference-in-differences design, I find evidence that opioid overdose deaths fell immediately following the first giveaway but increased following the second giveaway and discuss these apparently contradictory findings in the context of the changing composition of the opioid supply. I also find some evidence of a decline in opioid overdose-related ED visits following the giveaways. This study is the first to examine the effects of untargeted naloxone distribution and has implications for other novel, naloxone distribution efforts currently underway. [Abstract copyright: © 2023 The Authors. Health Economics published by John Wiley & Sons Ltd.]