The relationship between microvascular network resistance and functional capillary density was investigated by computer simulation. A model of spinotrapezius muscle microcirculatory network was constructed from measured data of vessel length and diameter. Progressive rarefaction was simulated by random stepwise removal of capillaries from the network. The flow distribution in the network was calculated after each step of capillary removal. Network resistance increased little when decreases in functional capillary density were less than 30%. Furthermore, capillary flow did not decrease and its variance across the network was unchanged despite the gradual flow redistribution. However, with further decreases in capillary density, network resistance increased in a double exponential fashion. The resistance approached infinity at a critical value of functional capillary density termed the percolation threshold. The value of this parameter varied between 56 and 62% and it was lower when capillaries with the lowest flow were preferentially removed.