Chronic bacterial infections of the lung are shaped by spatial heterogeneity across anatomical and cellular niches. Integrating deep whole-genome sequencing, metagenomics, and spatial transcriptomics, we have characterised within-host evolution and gene expression of key respiratory pathogens, including Pseudomonas aeruginosa. Spatially resolved and longitudinal sampling reveals within-patient translocation between body sites and identifies previously unrecognized niches within diseased lung tissue. By adapting spatial transcriptomics for bacterial detection, we find that bacterial behaviour and host immune responses vary by microenvironment. These findings underscore the importance of spatial context in infection dynamics and provide new insights into chronic lung disease and host–pathogen interactions.