A method for the assessment of neuromuscular function by the detection and classification of late wave activity, comprising: (i) pre-processing an ensemble of traces so as to attenuate noise, PFI and baseline disturbances; (ii) identifying time segments of late wave activity in the ensemble of traces; (iii) classifying the time segments of late wave activity into F-wave, A-wave and Repeater segments by exploiting the variable morphology of F-waves across the traces and the fixed morphology of A-waves and Repeater waves across the traces; and (iv) searching the traces in the vicinity of the F-wave trace segments on a trace-by-trace basis so as to identify the F-wave onset latency.