Chaos is a mechanism which can generate random-looking time series fro
m simple deterministic systems. We outline the idea of deterministic c
haos and discuss some biological examples where chaos is believed to p
lay a part. The benefits that biological systems may gain from exhibit
ing chaotic dynamics are discussed. Establishing whether a system exhi
bits chaotic dynamics by examining a time series consisting of observa
tions of a single variable is not an easy task. Several sophisticated
techniques are examined, including dimensional analysis, estimation of
Lyapunov exponents and nonlinear forecasting. There are many difficul
ties in applying these techniques; often the data requirements are lar
ge and noise in the data can give misleading results. We discuss the m
ethod of surrogate data, which can help avoid fallacious results.