In this paper, we present and discuss the methodology to monitor average residual stresses in aged high temperature unidirectional and woven graphite fiber/PMR-15 composites. The aging experiments were performed at 315 โข C for 1170 h in nitrogen and 1064 h in air, respectively. The proposed methodology was based on X-ray diffraction (XRD) strain measurements in embedded crystalline metallic inclusions. The aging experiments performed in nitrogen showed an increasing trend in the measured residual stresses from the embedded inclusions with aging time. In contrast, the aging experiments performed in air showed an increasing trend up to โผ200 h and then dropped by โผ40% for both unidirectional and woven composites. To extract the average stresses in the composite matrix from the average stresses in embedded inclusions as determined using XRD, the well-established Eshelby/Mori-Tanaka method was used. In addition, to address the effect of inclusion interactions, we also computed the average stresses in the embedded inclusions by taking into account inclusion interactions within the frame-work of the Eshelby/Mori-Tanaka method. It was shown that the interactions between the inclusions did not significantly disturb the average stresses in the inclusions, thereby validating the Eshelby/Mori-Tanaka method used in our analysis. Finally, we show that for an aging process that does not cause any damage to a composite matrix, a relatively simple computation based on the visco-elastic Eshelby/Mori-Tanaka method is sufficient to determine the average stress field in a composite matrix as compared with the predictions based on classical lamination plate theory (CLPT). However, if significant matrix degradation occurs, the current micromechanical modeling approach would likely produce erroneous results.