This study focused on the steady-state phase of exercise to evaluate the relative contribution of metabolic instability (measured with NIRS and haematochemical markers) and muscle activation (measured with EMG) to the oxygen consumption ($${\dot{V}O_2}$$
V
˙
O
2
) slow component ($${\dot{V}O_2}{_s}{_c}$$
V
˙
O
2
s
c
) in different intensity domains. We hypothesized that (i) after the transient phase, $${\dot{V}O_2}$$
V
˙
O
2
, metabolic instability and muscle activation tend to increase differently over time depending on the relative exercise intensity and (ii) the increase in $${\dot{V}O_2}{_s}{_c}$$
V
˙
O
2
s
c
is explained by a combination of metabolic instability and muscle activation. Eight active men performed a constant work rate trial of 9 min in the moderate, heavy and severe intensity domains. $${\dot{V}O_2}$$
V
˙
O
2
, root mean square by EMG (RMS), deoxyhaemoglobin by NIRS ([HHb]) and haematic markers of metabolic stability (i.e. [La−], pH, HCO3−) were measured. The physiological responses in different intensity domains were compared by two-way RM-ANOVA. The relationships between the increases of [HHb] and RMS with $${\dot{V}O_2}$$
V
˙
O
2
after the third min were compared by simple and multiple linear regressions. We found domain-dependent dynamics over time of $${\dot{V}O_2}$$
V
˙
O
2
, [HHb], RMS and the haematic markers of metabolic instability. After the transient phase, the rises in [HHb] and RMS showed medium–high correlations with the rise in $${\dot{V}O_2}$$
V
˙
O
2
([HHb] r = 0.68, p < 0.001; RMS r = 0.59, p = 0.002). Moreover, the multiple linear regression showed that both metabolic instability and muscle activation concurred to the $${\dot{V}O_2}{_s}{_c}$$
V
˙
O
2
s
c
(r = 0.75, [HHb] p = 0.005, RMS p = 0.042) with metabolic instability possibly having about threefold the relative weight compared to recruitment. Seventy-five percent of the dynamics of the $${\dot{V}O_2}{_s}{_c}$$
V
˙
O
2
s
c
was explained by [HHb] and RMS.