Irruptive bark beetles usually co-occur with their co-evolved tree hosts at very low (endemic) population densities. However, recent droughts and higher temperatures have promoted widespread tree mortality with consequences for forest carbon, fire and ecosystem services (Kurz et al., 2008; Raffa et al., 2008; Jenkins et al., 2012). In this issue of New Phytologist, Netherer et al. (pp. 1128–1141) experimentally explore the direct link between tree symptoms of drought and spruce bark beetle attack success rate. The study combined precipitation removal with a novel method for assessing bark beetle attacks. Lower soil moisture promoted lower tree water potentials, relatively lower tree resin flow, and a higher proportion of successful bark beetle attacks. Although attack rates were low, their results also suggest that host attractiveness to beetles decreased at the highest level of water stress. The Netherer et al. paper highlights the complex nature of interactions of trees with bark beetles. For example, the bark beetles show variability in the propensity to attack trees that may or may not be tied to environmental and tree cues. Factors related to the intrinsic beetle biology, combined with changes in tree physiology, highlight the difficulty of unraveling these interactions. In this commentary, we briefly review this complexity and offer suggestions for making further progress on this important problem particularly from the point of view of tree physiology.