Quantitative analysis of electromyographic (EMG) data still demands new methods. One of the methods is based on counting occurrences of signal crossings of "noise level" (per time unit). This provides an estimator of the socalled "aggregate activity". These data can be obtained by a spike trigger connected to a computer. Comparison of total activity of firing motor units in intact and affected muscles may give important information about neuromuscular system in norm and pathology. The aggregate activity method does not seem, however, to be completly objective, because its results depend on the value of the noise level selected by the experimenter. Thus, we propose the modification that allows: (1) quantitative estimation of how the results obtained with the aggregate activity method depend on the value of selected level, and (2) assessment of the distribution of the EMG signal amplitude -i.e., appearance particular height spikes in the signal. The advantages and disadvantages of the method are discussed in the context of two experimental models.