This paper employs equivalent circuit analysis of electrochemical impedance spectra (EIS) signatures of commercial proton exchange membrane fuel cell stacks (PEMFC) to establish a relation between oxygen concentration and EIS response within individual cells of the stack. This paper is part of an ongoing effort to use EIS for detecting and quantifying hydrogen crossover to understand and reduce hydrogen cathode exhaust emissions. Selected results from different sets of experiments performed on Ballard water-cooled and air-cooled PEMFC stacks are presented. A novel reduced-voltage EIS system with the ability to perform impedance measurements over different cell counts is employed to record EIS scans of these stacks. In the first set of experiments, hydrogen was injected into the cathode side of a 30 kW 150+-cell water-cooled module at various rates to emulate the phenomenon of hydrogen crossover at different levels. In the second set of experiments, nitrogen was injected into a target single-cell in an air-cooled stack with over 50 cells (50+-cell) to dilute the cathode available oxygen, and the impedance was measured over various numbers of cells spanning around this target cell. A novel equivalent circuit model with resistors, constant phase elements (CPEs), and a linear finite Warburg parameter was used to study the EIS data via an optimization-based fitting approach in python. Results of these experiments have shown there is a proportional relationship between the Warburg parameter (in the equivalent circuit model used) and hydrogen crossover (at nominal idle and low-current air flows), and that this Warburg parameter increases significantly with reduced cathode air flows. Also, it is concluded that this relationship can be further used, along with decomposition of the EIS scan of a stack into different cell-level scans, to estimate the number of large-leaky cells as well as the total hydrogen crossover rate.