Confinement controls bacterial spreading at all scales.

In natural environments such as soils or physiological ducts, swimming microorganisms alternate between three- and two-dimensional motion, exhibiting distinct kinematics. This interplay between three- and two-dimensional motion is intensified under confinement, yet its impact on long-range diffusivity remains poorly understood. Bacterial transport is also influenced by internal biochemical processes that determine tumbling rates and can produce long-term memory effects. Here, we combine long-duration three-dimensional tracking of wild-type Escherichia coli between parallel surfaces of separation H with a non-Markovian stochastic model to quantify how confinement regulates spreading. We measure diffusivity and bulk/surface residence times as a function of H and find excellent agreement with theory, with no need for parameter fitting. We derive an analytical expression for diffusivity, demonstrating that large-scale dispersion is governed by the average bulk residence time that can be obtained from a first-passage problem with two key limits: when H is smaller than, or larger than, the bacterial persistence length. This is the first experimental study to establish a direct link between confinement height and bacterial diffusivity. It provides a general framework for understanding how microbes navigate in geometrically complex environments.