(1) Pions produced in the development of extended atmospheric cosmic ray air showers subsequently decay to muons. The measured yield of those muons is generally underestimated by current phenomenological models and event generators optimized for cosmic ray physics. The importance of those disagreements motivates the feasibility studies for testing these models at the Large Hadron Collider (LHC) energies, at the highest center-of-mass energies achievable in a laboratory. The interaction of a nucleus and a virtual pion created in a charge exchange reaction at the LHC is a similar process to those contributing to the development of air showers in case of cosmic rays. The crucial problem of such an analysis is the selection of charge exchange events with the highest possible efficiency and high purity from proton–proton collisions at the LHC. (2) For this we consider distributions of various measurable quantities given by event generators commonly used in cosmic ray physics. (3) We examine the expected distributions of energy deposited in different calorimeters of an LHC experiment. We consider the geometrical acceptance and energy resolution of the detectors at the Compact Muon Solenoid (CMS) experiment, as an example. We determine a working point cut from the various options for event selection, and compare signal and background predictions using different models for a representative simple observable, such as average transverse momentum or charge particle yield. (4) A set of event selection cuts along these considerations is proposed, with the aim of achieving optimal efficiency and purity.