The advent of per call measurement data (PCMD) individual subscriber experiences and permits the construction of detailed sub-sector-level distribution maps. These maps can additionally be conditioned on the class of service (voice, data, short message services [SMS]), call conclusion status (normal, blocked, dropped), and other characteristics to yield valuable insight into network operation. As such, accurate geo-location techniques both are a valuable end in themselves and serve as an important starting point for future efforts involving performance improvements and even path-loss estimation (where subscribers will eventually replace the now-ubiquitous use of drive-test equipment in sufficiently mature markets).This paper provides an overview of per call measurement data and describes a maximum likelihood geo-location algorithm. We discuss an evaluation of this approach using drive test data. Examples of user density maps and CDMA market analyses are then provided, followed by our conclusions.