Without doubt the basic reproductive ratio, R-0, is the most widely used qu
antity in epidemic theory. Standard compartmental models show how R-0 is re
lated to the average age of infection, vaccination thresholds for eradicati
on and equilibrium solutions. However, many of the basic formulae for R-0 b
reak down when we consider transmission of infection to be a stochastic pro
cess involving discrete individuals. This paper clarifies why and when thes
e differences arise and predicts when individual-based considerations are l
ikely to be important in modelling infection dynamics. (C) 2000 Academic Pr
ess.