Support for the virulence trade-off theory is good but not perfect. Parasites differ in their degree of host generalism. Generalism is likely to impact virulence in four ways. Incorporation of infectivity data may lead to better predictions of evolved virulence.