The physical mechanism that drives bioirrigation is strongly dependent on the permeability of the sediment. We advance two mechanisms, each described by a corresponding microenvironment model. In muds, burrow water cannot penetrate the sediment, so bioirrigation is intrinsically driven by diffusional transfer across the burrow wall. This ‘‘diffusive’’ mode of bioirrigation is accurately described by the classical tube irrigation model. In sands, ventilation flows can penetrate the surrounding sediment via dead end burrows. To quantify this ‘‘advective’’ mode of bioirrigation, we propose a novel two-dimensional pocket injection model. This model’s principal features are that (1) organisms indent the sediment–water interface with burrow structures, (2) the specific structure of the burrow can be neglected except for the location of a feeding pocket, and (3) burrow water is injected from this feeding pocket into the surrounding sediment. We tested the adequacy of the pocket injection model in a detailed case study of the lugworm Arenicola marina, comparing model simulations and experimental data from core incubations. Simulation of two different sets of inert tracer experiments shows good agreement between model and data, indicating that our model captures the relevant aspects of lugworm bioirrigation in permeable sediments.