Constitutional Measurement Science (CMS-Ω) introduces a pre-empirical boundary science that decides when measurement outputs may coherently function as scientific assertions—specifically as claims of existence, comparison/ranking, or evaluative authority. CMS-Ω does not propose metrics, estimators, benchmarks, or experimental procedures. Instead, it supplies semantic-structural necessity conditions that must hold before measurement outputs can be elevated into truth-apt roles under irreversible epistemic use. The paper identifies the measurement fallacy: treating operationalization → measurement → existence → evaluability as automatically valid. CMS-Ω argues that measurement becomes a role-conferring act when its outputs are used to justify ontology, ranking, or authority, and that this elevation incurs semantic irreversibility (loss of counterfactuals, accountability, and meaning-space) that cannot be “corrected later” by improved instrumentation or statistics. The framework is explicitly negative-law: it returns binary role-coherence verdicts (admissible / inadmissible), not recommendations. CMS-Ω formalizes five canonical preconditions for role-admissible measurement: (i) ontological predefinition of what is measured (non-generativity), (ii) dimensional stability, (iii) compatibility with semantic irreversibility, (iv) witness feasibility under finite interpretive/accountability capacity γ, and (v) regime locality (no regime transfer without re-grounding). From these it derives invariants governing admissibility order (existence → measurement → estimation → evaluation), witness conservation, and dominance of semantic irreversibility. The paper further catalogs pre-statistical failure patterns (e.g., proxy substitution at genesis, dimensional laundering, semantic reversibility injection, regime smearing, aggregation-induced meaning collapse, witness erasure) and defines a strict conjunction-form admissibility condition CMS_OK(M). Finally, it specifies an adversarial falsifiability criterion: CMS-Ω is refuted only by sustained counterexamples where a role-inadmissible measurement nonetheless stably bears authoritative scientific roles across regime change and irreversible use without semantic collapse. Keywordsconstitutional measurement; measurement admissibility; role-coherence; pre-empirical boundary science; semantic irreversibility; witness capacity; accountability; regime locality; proxy measures; dimensional analysis; measurement fallacy; epistemic authority; pre-statistical failure modes; evaluation legitimacy; invariants; falsifiability; negative results