Background: In the collagen-induced arthritis (CIA) mouse model, inflammation readouts are usually quantified using operator-dependent clinical scoring systems, and no systematic relationship with functional deficits has been detected. In this study, we extensively quantified sensory and motor deficits in CIA mice during natural disease progression and therapeutic treatment. Then, we used these data to build a scale to predict functional deficits on the basis of the classical clinical score. Methods: Using the CIA mouse model, we longitudinally screened multiple approaches to assess locomotion (open field test, Catwalk™), sensitivity (Von Frey, Hargreaves, static weight-bearing tests), and inflammation (skin temperature), and identified the most accurate tests to correlate sensory and motor deficits with disease severity, measured by clinical score. We then used these tests to characterize functional deficits in control (naïve and mice injected with complete Freund’s adjuvant) and CIA mice, either untreated or treated with methotrexate to prevent functional deficits. By mathematical approaches, we finally investigated the relationship between functional deficits and clinical score. Results: We found that the functional disability scores obtained with the open field, Catwalk™, Hargreaves, and skin temperature tests significantly correlated with the clinical score in CIA mice, either untreated or treated with methotrexate. Mathematical correlation showed that motor deficits, robustly characterized by two different tests, were twice more responsive than thermal sensitivity deficits. Conclusion: We propose the arthritis sensory and motor (ArthriSM) scale as a new theranostic tool to predict motor and sensory deficit based on the clinical score, in the experimental mouse model of CIA. This ArthriSM scale may facilitate the transfer of knowledge between preclinical and clinical studies.