Five methods for estimating isometric recruitment curves of electrically stimulated muscle were tested on the quadriceps muscle of able-bodied human subjects. Two discrete methods, steady-state step response and peak impulse response, and three continuous methods were examined. The continuous methods were ramp deconvolution via an idealized impulse response, ramp deconvolution via a measured impulse response and ramp response shifted via correlation. The results showed that variation in estimates of recruitment curves were dependent on the method of measurement and means of post-processing. The step response method is the traditional procedure for estimating recruitment curves; however, the ramp methods all appeared to offer acceptable accuracy. The correlation method was simplest and fastest to process. The peak impulse method was less repeatable and often differed from the other methods in predicting the stimulation threshold.