The laboratory short encapsulation pull-out test (LSEPT) has been widely accepted as the most efficient method to characterize the mechanical behaviour of cable bolts under axial loading. In this study, a number of LSEPTs was performed on conventional (Plain SuperStrand) and modified (MW9S) cable bolts using the improved pull-out test design. The effects of several parameters including the uniaxial compressive strength (UCS) of the confining medium and grout and the borehole diameter on the mechanical behaviour of MW9S and Plain SuperStrand cable bolts were investigated. Response surface methodology (RSM) was employed to quantify the contribution of these parameters on the responses including peak and residual loads and initial stiffness. RSM revealed that the UCS of the confining medium is a key contributing factor to the mechanical behaviour of both cable bolts. Also, it was demonstrated that the borehole diameter had a negligible impact on the overall behaviour of the MW9S cable bolt while the peak load of the SuperStrand cable bolt was increased due to an increase in the diameter of the borehole. Finally, from a comparative analysis, it was confirmed that the improved design can better represent the field mechanical behaviour of cable bolts through maintenance of the embedment length during large deformation.