A number of mathematical theories to describe the momentum and energy
balances associated with running have been advocated since the pioneer
ing work of the British biologist A.V. Hill in the 1920s. Since the va
rious models lead to quite different conclusions and predictions about
the balances associated with running, it is instructive to examine cr
itically the assumptions underlying them. So, for example, depending o
n the model employed, it has been predicted that, for a sprinter, the
relative expenditure of energy in overcoming air resistance lies somew
here in the range of 2 to 16%. This wide range of values derives mainl
y from the various estimates of the propulsive force exerted during sp
rinting, indicating just how disparate the predictions among the model
s are. In the early 1970s Keller built on the ideas of Hill to develop
a complete theoretical model for competitive running in which the goa
l of the athlete is to cover the race distance in a minimum amount of
time (subject to certain constraints on the runner.) While the Hill-Ke
ller model provides only a first approximation, in that no attempt is
made to model separately the anaerobic and the aerobic pathways for en
ergy conversion and only the simplest model for the resistance to moti
on is considered, it nevertheless gives a good overall representation
of the world running records from 60 yds. to 10 km. Based on a careful
appraisal of the various models, it is suggested here that approximat
ely 3% of a sprinter's effort is spent on overcoming air resistance, a
figure that is supported by an independent estimate derived from the
long jump. An interesting consequence of the analysis is the feature t
hat the benefits of sprinting at altitude cannot be accounted for whol
ly by the reduced air density, suggesting that there may be some physi
ological benefit to sprinting al altitude. As a special application of
the estimate for the cost of overcoming air resistance, an appraisal
has been made of the series of 100 m races run at the U.S. Olympic Tri
als in 1988 in which Florence Griffith-Joyner set a new world record.
This record has been treated with great suspicion in the athletics com
munity because of a questionable wind-gauge reading. It is shown here
that the record was almost certainly wind aided, with a trailing wind
of 4 ms-1 or more.