396
Journal of Strength and Conditioning Research, 2006, 20(2), 396±399
q 2006 National Strength & Conditioning Association
THE EFFECT OF REST INTERVAL LENGTH ON BENCH
PRESS PERFORMANCE WITH HEAVY VS. LIGHT LOADS
JEFFREY M. WILLARDSON1 AND LEE N. BURKETT2
1Physical Education Department, Eastern Illinois University, Charleston, Illinois 61920; 2Department of Exercise
and Wellness, Arizona State University East, Mesa, Arizona 85212.
ABSTRACT. Willardson, J.M., and L.N. Burkett. The effect of rest
interval length on bench press performance with heavy vs. light
loads. J. Strength Cond. Res. 20(2):396–399. 2006.ÐThe purpose
of the current study was to compare the effect of 3 different rest
intervals on multiple sets of the bench press exercise performed
with heavy vs. light loads. Sixteen resistance-trained men per-
formed 2 testing sessions each week for 3 weeks. During the ®rst
testing session each week, 5 consecutive sets of the bench press
were performed with 80% of 1 repetition maximum (1RM) and
with a 1-, 2-, or 3-minute rest interval between sets. During the
second testing session each week the same procedures were re-
peated with 50% of 1RM. The total repetitions completed and
the sustainability of repetitions were compared between rest
conditions and between loads. For each load, resting 3 minutes
between sets resulted in signi®cantly greater total repetitions
vs. resting 2 minutes (p 5 0.000) or 1 minute (p 5 0.000) be-
tween sets. However, the sustainability of repetitions was not
signi®cantly different between loads (p 5 0.849). These results
can be applied to weekly bench press workouts that undulate
between heavy (i.e., 80% 1RM) and light (i.e., 50% 1RM) inten-
sities. When the training goal is maximal strength development,
3 minutes of rest should be taken between sets to avoid signif-
icant declines in repetitions. The ability to sustain repetitions
while keeping the intensity constant may result in a higher
training volume and consequently greater gains in muscular
strength.
KEY WORDS. recovery, repetitions, fatigue, strength, sets
INTRODUCTION
R
esistance training has been recognized as an
essential component of a comprehensive ®t-
ness program for individuals with diverse ®t-
ness goals. Individuals may participate in re-
sistance training for rehabilitative reasons or
in preparation for strenuous jobs such as ®re ®ghting,
law enforcement, or military service. Several training
variables must be considered when prescribing a resis-
tance exercise program.
Manipulation of training variables such as intensity,
volume, frequency, repetition velocity, and rest between
sets is dependent on the speci®c goals of the individual
and the physical demands faced during daily life (1, 2).
Although training variables such as intensity and volume
have been researched most frequently in determining
muscular adaptations, relatively little research has been
conducted on the optimal rest interval between sets.
The length of rest interval must be suf®cient to recov-
er energy sources (i.e., adenosine triphosphate [ATP] and
phosphocreatine), clear fatigue-producing substances
(i.e., hydrogen ions), and restore force production (3, 17).
Previous studies have demonstrated that the length of
the rest interval was an important factor affecting the
total repetitions that could be completed during a resis-
tance workout (4, 6, 9, 15, 18). Because of the popularity
of the bench press exercise, further research that exam-
ines rest intervals, speci®cally for this exercise, would be
useful for athletes to sustain repetitions without exces-
sive rest betweens sets. The ability to sustain repetitions
while keeping the intensity constant may result in a high-
er training volume and, consequently, greater gains in
muscular strength (11). Therefore, the purpose of the cur-
rent study was to compare the effect of 3 different rest
intervals on multiple sets of the bench press exercise per-
formed with heavy vs. light loads.
METHODS
Approach to the Problem
Data collection occurred over a period of 4 weeks. During
week 1, maximal strength for the bench press exercise
was determined for each subject on 2 different days using
standardized procedures (5, 12). The highest value of the
two 1 repetition maximum (1RM) measurements was
used to determine the load assignments for the other test-
ing sessions. During weeks 2, 3, and 4, subjects performed
2 testing sessions each week, the sessions being separated
by 72 hours, the ®rst with a load of 80% of 1RM and the
second with a load of 50% of 1RM. Five sets were per-
formed with each load to the point of voluntary exhaus-
tion. The rest interval between sets was 1, 2, or 3 minutes
counterbalanced. All subjects were required to perform 5
sets with each load and each rest condition one time,
making this study a within-design study.
Subjects
Sixteen men were used during the research process (age,
26.75 6 6.40 years; height, 178.75 6 6.00 cm; body mass,
92.10 6 19.61 kg). Maximal strength for the bench press
averaged 119.89 6 21.76 kg, and subjects were able to lift
an average of 1.34 6 0.29 times their body mass. All sub-
jects were considered recreationally trained lifters and
had practiced a body-building style of training that in-
volved multiple sets performed with light to moderate
loads (i.e., 50±80% maximal strength) combined with
short to moderate rest intervals between sets (i.e., 1±2
minutes).
To qualify for inclusion in the study, subjects had to
be apparently healthy, as determined by the Physical Ac-
tivity Readiness Questionnaire. Prior to data collection,
subjects were required to sign a consent form in accor-
dance with human subject regulations. Subjects were per-

RECOVERY 397
TABLE 1. Total repetitions 50% and 80% 1 repetition maximum.
Load Rest Minimum Maximum Mean SD Post Hocs Signi®cance
50% 1
2
3
45.00
53.00
65.00
75.00
94.00
105.00
59.13
74.81
87.69
10.31
12.36
13.15
1 vs. 2
1 vs. 3
2 vs. 3
0.000
0.000
0.000
80% 1
2
3
12.00
16.00
19.00
25.00
34.00
35.00
18.06
23.06
27.06
4.64
5.95
5.37
1 vs. 2
1 vs. 3
2 vs. 3
0.000
0.000
0.005
FIGURE 1. Performance patterns total repetitions. 1RM 5 1
repetition maximum.
TABLE 2. Post-hoc sets and rest intervals.
Sets Signi®cance Rest (minutes) Signi®cance
1 vs. 2
1 vs. 3
1 vs. 4
0.000
0.000
0.000
1 vs. 2
1 vs. 3
2 vs. 3
0.000
0.000
0.000
1 vs. 5
2 vs. 3
2 vs. 4
2 vs. 5
3 vs. 4
3 vs. 5
4 vs. 5
0.000
0.000
0.000
0.000
0.000
0.000
0.018
mitted to continue with their normal workouts through-
out the research period, with the following exceptions: (a)
subjects were asked not to perform the bench press other
than during their scheduled testing sessions, and (b) sub-
jects were asked not to work out on the same day before
their testing session.
Procedures
Warm-up sets were performed prior to each bench press
session. For the 50% test condition, subjects performed 1
warm-up set at 50% of the goal resistance for 10 repeti-
tions. For the 80% test condition, subjects performed 2
warm-up sets; the ®rst with 50% of the goal resistance
and the second with 75% of the goal resistance, for 10
repetitions each (6, 18). Subjects were instructed to per-
form repetitions with the intent to move the resistance as
rapidly as possible.
The bench press was performed with an Olympic bar
through the full range of motion. Subjects descended to
the point at which the resistance touched the chest, be-
fore pressing the resistance back to the starting point
with the elbows extended. One spotter was used during
all sets to assist in racking the resistance and to ensure
that subjects maintained consistent and safe technique
(i.e., hyperextending the lumbar spine or bouncing the
bar off the chest was not permitted). The rest interval
between sets was timed using a hand-held stopwatch.
Statistical Analyses
For each load, a 1-way repeated analysis of variance (AN-
OVA) was conducted to compare the total repetitions com-
pleted for each rest condition. Effect size, power, and lin-
ear and quadratic contrasts were calculated to assess the
practical importance of the length of the rest interval on
the total repetitions completed. The sustainability of rep-
etitions was then compared by dividing the number of
repetitions completed on each subsequent set by the num-
ber of repetitions completed on the ®rst set, with the re-
sultant percentages being used in a 2 (loads) by 3 (rest
intervals) by 5 (sets) repeated ANOVA. Linear and qua-
dratic contrasts were calculated to assess the perfor-
mance patterns associated with repeated sets to exhaus-
tion. An alpha level of 0.05 was used to determine signif-
icance for all comparisons.
RESULTS
For each load, a 1-way repeated ANOVA indicated that
the total repetitions completed were signi®cantly differ-
ent between rest conditions (p 5 0.000). Post-hocs were
then conducted using the Bonferroni adjustment to de-
termine speci®c differences (Table 1). The effect size sta-
tistics for the 50% and 80% loads indicated that 93.5%
and 78.2% of the variance in the total repetitions com-
pleted could be accounted for by the length of the rest
interval. For each load, the power statistic was 1.000,
which indicated that these results would occur 1,000
times if this experimental design were tested on the same
sample size. For each load, a signi®cant linear relation-
ship existed between the length of the rest interval and
the total repetitions completed (p 5 0.000; see Figure 1).
The 2 (loads) by 3 (rest intervals) by 5 (sets) repeated
ANOVA indicated that the sustainability of repetitions
was signi®cantly different between sets (p 5 0.000) and
between rest conditions (p 5 0.000) but was not signi®-
cantly different between loads (p 5 0.849). Post-hocs were
then conducted using the Bonferroni adjustment to de-
termine speci®c differences between sets and between
rest conditions (see Table 2). Signi®cant linear (p 5
0.000) and quadratic (p 5 0.000) contrasts indicated that
a large decline in repetitions occurred between the ®rst
set and the third set, followed by a small decline between
the third set and the ®fth set (Figure 2 and Table 3).
Since the sustainability of repetitions was not signi®-
cantly different between loads, Figure 2 represents the
mean percentages of both loads combined.