There is a need for easy-to-use methods to assess training progress in sports and rehabilitation research. The present review investigated whether cardiac acceleration at the onset of physical exercise (HRonset) can be used as a monitoring variable. The digital databases of Scopus and PubMed were searched to retrieve studies investigating HRonset. In total 652 studies were retrieved. These articles were then classified as having emphasis on HRonset in a sports or rehabilitation setting, which resulted in 8 of 112 studies with a sports application and 6 of 68 studies with a rehabilitation application that met inclusion criteria. Two co-existing mechanisms underlie HRonset: feedforward (central command) and feedback (mechanoreflex, metaboreflex, baroreflex) control. A number of studies investigated HRonset during the first few seconds of exercise (HRonsetshort), in which central command and the mechanoreflex determine vagal withdrawal, the major mechanism by which heart rate (HR) increases. In subsequent sports and rehabilitation studies, interest focused on HRonset during dynamic exercise over a longer period of time (HRonsetlong). Central command, mechanoreflexes, baroreflexes, and possibly metaboreflexes contribute to HRonset during the first seconds and minutes of exercise, which in turn leads to further vagal withdrawal and an increase in sympathetic activity. HRonset has been described as the increase in HR compared with resting state (delta HR) or by exponential modeling, with measurement intervals ranging from 0–4 s up to 2 min. Delta HR was used to evaluate HRonsetshort over the first 4 s of exercise, as well as for analyzing HRonsetlong. In exponential modeling, the HR response to dynamic exercise is biphasic, consisting of fast (parasympathetic, 0–10 s) and slow (sympathetic, 1–4 min) components. Although available studies differed largely in measurement protocols, cross-sectional and longitudinal training studies showed that studies analyzing HRonset in relation to physical training primarily incorporated HRonsetlong. HRonsetlong slowed in athletes as well as in patients with a coronary disease, who have a relatively fast HRonsetlong. It is advised to include both HRonsetlong and HRonsetshort in further studies. The findings of this review suggest that HRonset is a potential tool for monitoring and titrating training in sports as well as in rehabilitation settings, particularly in patients with ventricular fibrillation. Monitoring HRonset in the early phase of training can help optimize the effectiveness of training and therapy. More research is needed to gain a better understanding of the mechanisms underlying HRonset in relation to their application in sports and rehabilitation settings.