The combination of frequency-tagging with electroencephalography (EEG) has recently proved fruitful for understanding the perception of beat and meter in musical rhythm, a common behavior shared by humans of all cultures. EEG frequency-tagging allows the objective measurement of input–output transforms to investigate beat perception, its modulation by exogenous and endogenous factors, development, and neural basis. Recent doubt has been raised about the validity of comparing frequency-domain representations of auditory rhythmic stimuli and corresponding EEG responses, assuming that it implies a one-to-one mapping between the envelope of the rhythmic input and the neural output, and that it neglects the sensitivity of frequency-domain representations to acoustic features making up the rhythms. Here we argue that these elements actually reinforce the strengths of the approach. The obvious fact that acoustic features influence the frequency spectrum of the sound envelope precisely justifies taking into consideration the sounds used to generate a beat percept for interpreting neural responses to auditory rhythms. Most importantly, the many-to-one relationship between rhythmic input and perceived beat actually validates an approach that objectively measures the input–output transforms underlying the perceptual categorization of rhythmic inputs. Hence, provided that a number of potential pitfalls and fallacies are avoided, EEG frequency-tagging to study input–output relationships appears valuable for understanding rhythm perception.