In contrast to the large mass, complexity and recirculating power of conventional drivers for inertial confinement fusion (ICF), antiproton annihilation offers a specific energy of 90 MJ µg−1 and thus a unique form of energy packaging and delivery. In principle, antiproton drivers could provide a profound reduction in system mass for advanced space propulsion by ICF. We examine the physics underlying the use of antiprotons () to drive various classes of high-yield ICF targets by the methods of volumetric ignition, hotspot ignition and fast ignition. The useable fraction of annihilation deposition energy is determined for both -driven ablative compression and -driven fast ignition, in association with zero- and one-dimensional target burn models. Thereby, we deduce scaling laws for the number of injected antiprotons required per capsule, together with timing and focal spot requirements. The kinetic energy of the injected antiproton beam required to penetrate to the desired annihilation point is always small relative to the deposited annihilation energy. We show that heavy metal seeding of the fuel and/or ablator is required to optimize local deposition of annihilation energy and determine that a minimum of ∼3 × 1015 injected antiprotons will be required to achieve high yield (several hundred megajoules) in any target configuration. Target gains—i.e. fusion yields divided by the available p– annihilation energy from the injected antiprotons ()—range from ∼3 for volumetric ignition targets to ∼600 for fast ignition targets. Antiproton-driven ICF is a speculative concept, and the handling of antiprotons and their required injection precision—temporally and spatially—will present significant technical challenges. The storage and manipulation of low-energy antiprotons, particularly in the form of antihydrogen, is a science in its infancy and a large scale-up of antiproton production over present supply methods would be required to embark on a serious R&D programme for this application.