This paper summarizes the results of a series of experiments establish
ed to examine the effects of the composition and pattern of physical w
orkload on human heat tolerance and to compare the predictive abilitie
s of a selection of prominent heat stress indices. The experiments wer
e carried out in a climatically controlled chamber and involved mixtur
es of dynamic (treadmill walking) and static (weight holding) activiti
es as appropriate. The average physiological responses to intermittent
dynamic workloads did not significantly differ from those to continuo
us dynamic work of the same average intensity. However, the average va
lues masked periods of peak physiological loading. The inclusion of a
static load component had a negligible effect on the average level of
physiological thermal strain. As a consequence, the presence or absenc
e of a static work component did not influence the predictive abilitie
s of the five heat stress indices examined. It was concluded that, whe
re peak loading was unlikely to be a problem, time weighted averaging
of workload was a reasonable approach in assessing industrial thermal
strain and that heat stress criteria, derived originally for continuou
s dynamic workloads, were likely to be equally appropriate for intermi
ttent combined workloads. However, care should be taken to ensure that
peak loading did not present a health risk to industrial populations
carrying out intermittent work in hot environments.