In this paper we present tensile and 2-point bending strength results on 5 cm. (2-inch) sections of mechanically stripped fiber. We show that the ‘weakest-link’ model is apparently not obeyed since the low strength mode of the resulting distributions have different widths, i.e., the Weibull m-values for the tensile and 2-point tests are ~2 and ~7, respectively. This simply means that the two measurements are sampling different flaw populations. The importance of understanding this behavior and applying that understanding in practice is emphasized.