At present, one of the most problematic topics associated with the application of optical pyrometry to in-service use on gas turbine aeroengines is the fouling of the system optics. Even though only one optical surface is, in general, exposed to the turbine environment the resulting deposition of particulates, from the combustion process or atmospheric ingestion, can greatly inhibit the benefits of utilising such an instrument for direct temperature measurement.The particulates that deposit on the pyrometer lens act as an interference filter by absorbing a portion of the thermal radiation from the target, for example the turbine blades. As a control function input, this will then bias the indicated temperature low and thus permit higher turbine temperature operation resulting in blade temperatures in excess of their intended limits. In practice, a purge air system is therefore incorporated into turbine pyrometer designs in order to minimise the number of particulates that can potentially deposit and build-up on the exposed system optics.This paper examines and compares the flow fields in two purge air designs that have already gained acceptance inservice, namely the military RB199 and civil GE90 pyrometer purging systems. This study of the airflow within actual purge designs will allow gas turbine engineers develop an understanding of the fluid flow structures and how they presently impede the achievement of sufficient lens cleanliness.