The single fiber pull-out test was employed to experimentally model the failure of an embedded optical fiber in neat resin. The objective was to evaluate load transfer between resin and fiber through the evaluation of the sensitivity of the single fiber pull-out test to the physical parameters of the polymer matrix. This was accomplished first by appropriately interpreting the load versus extension trace, second by determining the effect of fiber coating and embedding resin on the load carrying ability of the single fiber composite, third by characterizing the fiber/coating/resin system with respect to the physical parameters of the polymer and the failure mechanism of the composite as the cure temperature was varied, and fourth, by correlating the independent parameters of the polymer and fracture data. For the first time, the load versus extension trace generated by such experiments was thoroughly interpreted and mathematically modeled. To this end, the embedding resin was physically characterized through the determination of the glass transition temperature, T_g, the relative change in volume with sample preparation and thus the resulting normal pressure exerted on the embedded fiber by the resin material. The experimental fracture data was quantified by determining the strain energy release rate, SERR, for initiation of crack propagation and, with the consideration of friction, its continuation, as well as the interfacial shear stress, τ, of the bond, and τ associated with debonding and sliding. Based on a series of experiments of varying material parameters, a model material system was chosen: a polyimide coated fiber embedded in uncatalyzed tetraglycidyl-4-4'-diaminodiphenylmethane with 4,4’-diaminodiphenylsulfone. Cure temperatures, T_cure, of 150, 177, 230 and 250°C were employed. The average critical strain energy release rates increased from the 150 to 177 to 230°C sample sets, then decreased for the 250°C sample set. Since the T_g of the fully cured resin is 260°C, these results support the hypothesis of increasing residual stress as a function of T_cure for cure in the vitreous state. In regards to the 250°C cure data set, since T_cure was within T_g - 30°C the internal pressures due to crosslinking were minimized due to cure in a rubber-like state. The residual pressure, independently determined from both the resin characterization and fracture data, increased by a factor of 2.4 with a temperature increase from 150 to 230°C for the two hour cure period. The strain energy release rate and sliding interfacial shear stress of pull-out increased by a factor of 2.54 and 2.1, respectively. The coefficient of friction remained statistically constant at 0.6. Based on this work, it is concluded that the single fiber pull-out test is sensitive to fiber/matrix interactions via the physical parameters of the material system. Also, the failure response of the single fiber composite can be predicted for well characterized matrices. / Ph. D.