Honey bees, Apis mellifera, of European origin are major pollinators of crops and wild flora. Their endemic and exported populations are threatened by a variety of abiotic and biotic factors. Among the latter, the ectoparasitic mite Varroa destructor is the most important single cause behind colony mortality. The selection of mite resistance in honey bee populations has been deemed a more sustainable solution to its control than varroacidal treatments. Because natural selection has led to the survival of some European and African honey bee populations to V. destructor infestations, harnessing its principles has recently been highlighted as a more efficient way to provide honey bee lineages that survive infestations when compared with conventional selection on resistance traits against the parasite. However, the challenges and drawbacks of harnessing natural selection to solve the varroa problem have only been minimally addressed. We argue that failing to consider these issues could lead to counterproductive results, such as increased mite virulence, loss of genetic diversity reducing host resilience, population collapses or poor acceptance by beekeepers. Therefore, it appears timely to evaluate the prospects for the success of such programmes and the qualities of the populations obtained. After reviewing the approaches proposed in the literature and their outcomes, we consider their advantages and drawbacks and propose perspectives to overcome their limitations. In these considerations, we not only reflect on the theoretical aspects of host-parasite relationships but also on the currently largely neglected practical constraints, that is, the requirements for productive beekeeping, conservation or rewilding objectives. To optimize natural selection-based programmes towards these objectives, we suggest designs based on a combination of nature-driven phenotypic differentiation and human-directed selection of traits. Such a dual strategy aims at allowing field-realistic evolutionary approaches towards the survival of V. destructor infestations and the improvement of honey bee health.