While it is widely recognized that treatment choices of health care providers vary across similar patients, reasons for this remain poorly understood. This paper estimates the spatial evolution of opioid prescription leniency of health care providers. Using exits and entries of primary care providers into local markets I document spillover effect in opioid prescription leniency across primary care practices. My results imply that an increase in opioid leniency of 1 standard deviation of a random provider peer increases the leniency of the focal provider by 7% of a standard deviation. Finally, I apply the network model to estimate how increased opioid use is harmful to patients’ labor market outcomes.