The results presented are of work on carbon/thermoplastic systems aimed at identifying the relevant mechanisms of micromechanical interaction in single filament composites (SFC) and improving the understanding of the relative role of the resin and the fiber in these mechanisms. The experimental strategy has included the study of SFC specimens deformed in either axial or transverse tension, and the complementary nature of these testing modes is highlighted. Theoretical shortcomings of the Kelly-Tyson model, in its Weibull representation, are discussed on the basis of observed mechanisms for formation of saturated fragments. It is argued that failure conditions of bare fibers incompletely describe those of pre-saturated imbedded fragments. On this basis two procedures are proposed for estimating the interphase shear transmissibility which, as opposed to the Kelly-Tyson method, do not require information about the tensile strength of the filament. Interphase shear transmissibilities estimated by these methods are compared and differences are rationalized on the basis of the observed damage mechanisms. These mechanisms are discussed in terms of the interplay between matrix properties such as polymer chain mobility, toughness and stiffness and the shear transmissibility of the interphase. The short beam shear strength of J2/carbon composites correlates very well with interphase shear transmissibility estimated with SFCs. Methods are proposed for the use of filament fragmentation curves which depict the average fragment aspect ratio versus the applied strain in a fragmentation test. Results show that the filament strain in these thermoplastic SFC lags behind the matrix strain. How these curves readily determine the relative effects of surface treatments on filament tensile strength and interphase shear transmissibility without resorting to specific information about filament strength is demonstrated. The curves are also useful for determining the fiber strength utilization effectiveness of a given resin.