Abstract Clinical trials on dentine hypersensitivity have been numerous and protocols varied. To date there is little consensus as to the conduct of studies on this poorly-understood yet common and painful dental condition. A committee of interested persons from academia and industry was convened to discuss the subject of clinical trials on dentine hypersensitivity and a consensus report is presented. A double-blind randomized parallel groups design is recommended, although cross-over designs may be used for the preliminary screening of agents. Subjects may have multiple sites scored. Sample size will be determined by estimating the variability in the study population, the effect to be detected and the power of the statistical test to be used. Subject selection is based on a clinical diagnosis of dentine hypersensitivity, excluding those with conflicting characteristics such as currently-active medical or dental therapy. The vestibular surfaces of incisors, cuspids and bicuspids are preferred as sites to be tested. A range of sensitivity levels should be included. Tactile, cold and evaporative air stimuli should be applied. Negative and benchmark controls should be incorporated. Most trials should last 8 weeks. Sensitivity may be assessed either in terms of the stimulus intensity required to evoke pain or the subjective evaluation of pain produced by a stimulus using a visual analog or other appropriate scale. The subject's overall assessment may be determined by questionnaire. Outcomes should be expressed in terms of clinically significant changes in symptoms. Follow-up evaluation is required to determine the persistence of changes. At least 2 independent trials should be conducted before a product receives approval.