Summary Predators can directly and indirectly influence host–parasite interactions by consuming infected individuals, by removing infectious parasite stages and by changing host traits (e.g. behaviour). Because such effects can affect infection positively or negatively, understanding the net effects of predation on pathogen transmission under natural conditions is important. We conducted a mesocosm experiment to examine the effects of predators on interactions between tadpole hosts (Pseudacris regilla) and trematode parasites (Ribeiroia ondatrae). We manipulated the presence of (non‐lethal, i.e., caged) predators of tadpoles (dragonfly larvae) and (potentially lethal) parasite predators (damselfly larvae) to evaluate their individual and combined effects on host infection. We expected that dragonflies would reduce tadpole activity and thereby increase parasite infection through a reduction in antiparasite behaviour. Because damselflies can consume parasites in the laboratory, we predicted that damselflies would lower infection by consuming parasites before they infected tadpoles. Our goal was to evaluate the net consequences of these predator‐mediated effects for host/prey infection. The presence of caged dragonflies reduced tadpole activity, resulting in a ˜50% increase in average infection load compared to treatments without predators. In contrast to our prediction that damselflies would reduce infection, damselflies elicited behavioural and morphological changes in hosts similar to dragonflies, with a comparable increase in parasite transmission. Thus, predator‐mediated effects were evident predominantly through changes in host/prey behaviour, rather than through changes in the abundance of parasites. The lack of a direct effect of predators on infection (i.e. via consumption of parasites) could be the result of the presence of alternative prey (zooplankton) or a mismatch in timing between visual predators feeding during the day and parasites released from the first intermediate host and infecting amphibians at night. The presence of predators also stimulated morphological defences in their tadpole prey, including increased tail and body depth. Interestingly, we found that parasite infection also induced morphological changes in tadpole tail and body depth, similar to changes produced by (non‐lethal) cues from predators. Parasites caused malformations in tadpoles, but there were no effects on tadpole growth or development from either parasites or predators. This research has key implications for linking predation and infectious disease in aquatic ecosystems. Our results emphasise the importance of indirect effects of predators on infection and highlight possible trade‐offs in mitigating the concurrent risks of predation and disease. Parasites can also alter host morphology through trait‐mediated effects similar to predators, supporting a broader inclusion of parasites in the study of the ecology of natural enemies.