Electrocardiogram RR intervals duration reduce rapidly in the first seconds of dynamic exercise mainly due to a cardiac vagal withdrawal. However, it remains unclear if this response varies between exercises performed with different body segments (i.e. arm vs. leg). Our aim was to compare the vagal withdrawal at the onset of arm and leg dynamic exercise. Cardiac vagal withdrawal was assessed by the 4-s exercise test (4sET), a pharmacologically validated and highly reliable procedure. Initially, 60 healthy subjects performed the 4sET using arms (ARM 4sET) and legs (LEG 4sET), in a random order. Later, 20 of them repeated the testing, controlling for cycling rate and time of exercise onset within a RR interval, potential intervenient variables. Similar results were found for ARM and LEG 4sET considering the RR interval duration obtained immediately before the onset of exercise (RRB) (mean +/- SEM; 852 +/- 23 vs. 857 +/- 23 ms; P = 0.55), the shortest RR interval duration during exercise (RRC) (570 +/- 10 vs. 563 +/- 10 ms; P = 0.22) and the cardiac vagal index (CVI), which is the ratio between RRB and RRC (1.49 +/- 0.03 vs. 1.52 +/- 0.03; P = 0.10). Furthermore, high intraclass correlation coefficients were found (RRB r (i) = 0.94, P < 0.001; RRC r (i) = 0.85, P < 0.001; CVI r (i) = 0.81, P < 0.001). We conclude that similar cardiac vagal withdrawal was induced by 4-s fast unloaded cycling exercise performed with arms or legs.