OBJECTIVE The overall aim of this study was to determine whether any changes occur in the pulpal structure of human primary teeth in association with physiological root resorption. METHODS The experimental material comprised 64 sound primary molars, obtained from children requiring routine dental extractions under general anaesthesia. Pulp sections were processed for indirect immunofluorescence using combinations of: (i) protein gene product 9.5 (a general neuronal marker); (ii) leucocyte common antigen CD45 (a general immune cell marker); and (iii) Ulex europaeus I lectin (a marker of vascular endothelium). Image analysis was then used to determine the percentage area of staining for each label within both the pulp horn and mid-coronal region. Following measurement of the greatest degree of root resorption in each sample, teeth were subdivided into three groups: those with physiological resorption involving less than one-third, one-third to two-thirds, and more than two-thirds of their root length. RESULTS Wide variation was evident between different tooth samples with some resorbed teeth showing marked changes in pulpal histology. Decreased innervation density, increased immune cell accumulation, and increased vascularity were evident in some teeth with advanced root resorption. Analysis of pooled data, however, did not reveal any significant differences in mean percentage area of staining for any of these variables according to the three root resorption subgroups (P > 0.05, analysis of variance on transformed data). CONCLUSIONS This investigation has revealed some changes in pulpal status of human primary teeth with physiological root resorption. These were not, however, as profound as one may have anticipated. It is therefore speculated that teeth could retain the potential for sensation, healing, and repair until advanced stages of root resorption.