Reaction–diffusion models were applied to gain insight into the aspects of biofilm infection and persistence by comparing mathematical simulations with the experimental data from varied bacterial biofilms. These comparisons, including three in vitro systems and two clinical investigations of specimens examined ex vivo, underscored the central importance of concentration gradients of metabolic substrates and the resulting physiological heterogeneity of the microorganisms. Relatively simple one-dimensional and two-dimensional (2D) models captured the:(1) experimentally determined distribution of specific growth rates measured in Pseudomonas aeruginosa cells within sputum from cystic fibrosis patients;(2) pattern of relative growth rate within aggregates of streptococcal biofilm harboured in an endocarditis vegetation;(3) incomplete penetration of oxygen into a Pseudomonas aeruginosa biofilm under conditions of exposure to ambient air and also pure oxygen;(4) localisation of anabolic activity around the periphery of P. aeruginosa cell clusters formed in a flow cell and attribution of this pattern to iron limitation;(5) very low specific growth rates, as small as 0.025 h− 1, in the interior of cell clusters within a Klebsiella pneumoniae biofilm in a complex 2D domain of variable cell density.