Motor cortex stimulus-response (S/R) curves are an indication of cortical excitability and are of relevance to topographical mapping. The aims of this study were to compare two different methods of collecting data to construct a S/R curve for transcranial magnetic stimulation (TMS) in the upper limbs, to identify reliable summary statistics for the S/R curve, and to determine whether S/R curves predicted motor threshold. Motor evoked potentials (MEP) were obtained from biceps brachii (BB) and abductor pollicis brevis (APB) muscles at rest with a circular coil centered at the vertex. Motor threshold was determined using a validated protocol. MEPs were obtained with 1% increments in intensity or from the average of five trials at 5% increments. The S/R relationships were fitted to linear, S-shaped, and Boltzmann functions. A linear function determined from the average of five trials accurately summarized our data (r2 from 0.6 to 0.9, P < 0.05, n = 8, for right APB and from 0.6 to 0.9, P < 0.05, n = 8 for right BB). The X-axis intercept of the line determined using these methods fell between the upper and lower limits of motor threshold in all eight subjects. We propose that MEP values obtained at intervals of 5% averaged over five trials, fitted to a linear function provides a practical means of assessing the S/R characteristics of TMS for proximal and distal upper limb muscles.