The lack of a testable model explaining how ventilation is regulated in different exercise conditions has been repeatedly acknowledged in the field of exercise physiology. Yet, this issue contrasts with the abundance of insightful findings produced over the last century and calls for the adoption of new integrative perspectives. In this review, we provide a methodological approach supporting the importance of producing a set of evidence by evaluating different studies together-especially those conducted in 'real' exercise conditions-instead of single studies separately. We show how the collective assessment of findings from three domains and three levels of observation support the development of a simple model of ventilatory control which proves to be effective in different exercise protocols, populations and experimental interventions. The main feature of the model is the differential control of respiratory frequency (fR) and tidal volume (VT); fR is primarily modulated by central command (especially during high-intensity exercise) and muscle afferent feedback (especially during moderate exercise) whereas VT by metabolic inputs. Furthermore, VT appears to be fine-tuned based on fR levels to match alveolar ventilation with metabolic requirements in different intensity domains, and even at a breath-by-breath level. This model reconciles the classical neuro-humoral theory with apparently contrasting findings by leveraging on the emerging control properties of the behavioural (i.e. fR) and metabolic (i.e. VT) components of minute ventilation. The integrative approach presented is expected to help in the design and interpretation of future studies on the control of fR and VT during exercise.