Behavioral ecologists might often benefit by the ability to directly m
easure an animal's body condition as an estimate of foraging success,
and ultimately fitness. Here we compare the reliability and effectiven
ess of three indices of body condition that have been heavily used in
the morphometrics literature. We examined the ratio index (body mass/b
ody size), the slope-adjusted ratio index (based on regression slopes
generated from a reference population), and the residual index (the re
siduals of a regression of body mass on body size). We present the res
ults of tests performed in the field and laboratory on two ecologicall
y and evolutionarily divergent spider species: the vagrant wolf spider
Pardosa milvina (Araneae, Lycosidae), and the colonial orb-weaver Met
epeira incrassata (Araneae, Araneidae). The ratio index correlated wit
h body size, which weakened the strength of conclusions that could be
drawn. The slope-adjusted ratio index requires an independent and larg
e data set with which to generate the expected values, and was likewis
e sensitive to body size. The residual index, with appropriate transfo
rmations to achieve homoscedasticity, was the most reliable index beca
use it did not vary with body size, and we recommend its general use i
n behavioral studies that require a condition estimate.