Little is known about current practices in high-value care (HVC) bedside teaching. A lack of instruments for measuring bedside HVC behaviors confounds efforts to assess the impact of curricular interventions. The authors aimed to define observable HVC concepts by developing an instrument to measure the content and frequency of HVC discussions.The authors developed the HVC Rounding Tool in four iterative phases, using Messick's validity framework. Phases 1 and 2 were designed to collect evidence of content validity, Phases 3 and 4 to collect evidence of response process and internal structure. Phase 1 identified HVC topics within the literature. Phase 2 used a modified Delphi approach for construct definition and tool development. Through two rounds, the Delphi panel narrowed 16 HVC topics to 11 observable items, categorized into three domains (quality, cost, and patient values). Phase 3 involved rater training and creation of a codebook. Phase 4 involved three iterations of instrument piloting. Six trained raters, in pairs, observed bedside rounds during 148 patient encounters in 2016. Weighted kappas for each domain demonstrated improvement from the first to third iteration: Quality increased from 0.65 (95% CI 0.55-0.79) to 1.00, cost from 0.58 (95% CI 0.4-0.75) to 0.96 (95% CI 0.80-1.00), and patient values from 0.41 (95% CI 0.19-0.68) to 1.00. Percent positive agreement for all domains improved from 65.3% to 98.1%. This tool, the first with established validity evidence, addresses an important educational gap for measuring the translation of HVC from theoretical knowledge to bedside practice.