Stress has been shown to affect perceptual processing and decision mak
ing in various domains. Two experiments examined the effect of stress
on compliance behavior. The first experiment also examined the effect
of warning location. In Experiment 1, participants were assigned rando
mly to one of four conditions in a 2 (stress)x2 (warning placement) be
tween-subjects design. Participants performed a chemistry task in whic
h they weighed and measured various chemical substances that appeared
potentially hazardous, but were actually safe. In the high stress cond
ition, participants were given a time limit to complete the task, and
the experimenter stood immediately adjacent to the participant, appear
ing to be monitoring the participant's performance. In the low stress
condition, participants were given as much time as they needed to comp
lete the task and the experimenter stood at a distance, out of the par
ticipant's field of view. A warning to wear mask and gloves was presen
t in one of two locations, either as a posted sign or placed within a
set of task instructions. Compliance with the warning (wearing of prot
ective equipment) was significantly higher among participants who were
under low stress and exposed to the within-instructions warning. In E
xperiment 2, the stress manipulation was separated into two factors: t
ime pressure (absence vs presence) and social monitoring (absence vs p
resence). Results showed that time pressure significantly reduced comp
liance compared with its absence, but social monitoring produced a sma
ll but non-significant compliance enhancement. The results point to th
e need to consider external factors in warning systems. (C) 1998 Elsev
ier Science Ltd. All rights reserved.