The effect of fluctuating environmental conditions (i.e. environmental stochasticity) on the evolution of virulence has been broadly overlooked, presumably due to a lack of connection between the fields of evolutionary epidemiology and insect ecology. Practitioners of the latter have known for a long time that stochastic environmental variations can impact the population dynamics of many insects, some of which are vectors of infectious diseases. Here we investigate whether environmental stochasticity affecting a vector’s life history can have an indirect impact on the evolutionarily expected virulence of the parasite, using Chagas disease as an example. We model the evolution of virulence using the adaptive dynamics framework, showing that parasite virulence should decrease when the vector’s dynamics randomly change in time. The decrease is even more pronounced when environmental variations are frequent and ample. This decrease in virulence can be viewed as a bet-hedging strategy: when a parasite is at a risk of not being transmitted (e.g. because vectors are scarce), its best option is to stay in the host longer – that is, to be less virulent. Lowering virulence is thus very similar to increasing iteroparity, a well-known risk-spreading strategy, and should be expected to evolve whenever parasite transmission varies randomly in time.