Data from Critical Heat Flux (CHF) and Critical Power (CP) tests are used for developing CHF and CP (or dryout) correlations to predict the maximum allowable heat flux at points of operational interest. CHF and CP tests are conducted using heater rod arrays representing fuel rods in a partial fuel bundle. Heat is directly generated in these heater rods depicting the axial power distributions of interest. The test rod array is housed in a slender vertical channel of metallic walls that are generally unheated. During testing, part of the heat generated in the heater rods is lost to the environment through the channel walls. Heat loss through the channel wall has the potential to favor channel thermal-hydraulic (T-H) conditions, and thereby make power measurements non-conservative. Typically, a single conservative estimate of the heat loss obtained from a separate isothermal test of the same test configuration is used regardless of the reactor conditions tested during a test campaign. This paper investigates the basis for the use of a single value for channel heat loss under varying reactor conditions. Results indicate that this practice has sound bases when applied with appropriate conservatisms.