A great deal of published experimental evidence shows that parasite-induced changes in host choice, and biting and survival rates of bloodsucking insects are an important source of bias in estimates of Vectorial Capacity. I argue here that, because of the difficulties of handling these and other biases in practice, indices (based on the Vectorial Capacity) which purport to measure transmission rate are best used in a comparative rather than an absolute way. They should attempt to measure, not transmission rate, but changes in transmission rate. In this case, the less restrictive assumption that parasites are responsible for constant bias (rather than no bias) leads to relatively small errors. The results encourage a more optimistic view of the performance of classical entomological indices such as the parous rate. They should also help to redirect the principal epidemiological question about vector–parasite interactions: future studies should be concerned with the impact of parasites, not directly on Vectorial Capacity, but on the comparative measures which are actually used to evaluate transmission rate.