PURPOSE: The Japan Prosthodontic Society developed a multi-axis assessment protocol to evaluate the complex variations in patients who need prosthodontic care, and to classify the level of treatment difficulty. A previous report found the protocol to be sufficiently reliable. The purpose of this multi-center cohort study was to evaluate the validity of this multi-axis assessment protocol. METHODS: The treatment difficulty was evaluated using the multi-axis assessment protocol before starting prosthodontic treatment. The time required for active prosthodontic treatment, medical resources such as treatment cost, and changes in the oral health-related QOL before and after treatment, were evaluated after treatment completion. The construct validity of this protocol was assessed by the correlation between the dentist's pre-operative subjective assessment of the treatment difficulty, and the level of difficulty determined by this protocol. The predictive validity was assessed estimating the correlations between a "comprehensive level of treatment difficulty" based on the four axes of this protocol and total treatment cost, total treatment time, and changes in the oral health-related QOL before and after treatment. RESULTS: The construct validity of this protocol was well documented except for psychological assessment. Regarding the predictive validity, the comprehensive level of treatment difficulty assessed before treatment was significantly correlated with the three surrogate endpoints known to be related to the treatment difficulty (total treatment cost, treatment time, and improvement in the oral health-related QOL). To further clarify the validity of the protocol according to patients' oral condition, a subgroup analysis by defects was performed. Analyses revealed that treatment difficulty assessment before treatment was significantly related to one or two surrogate endpoints in the fully edentulous patients and the partially edentulous patients. No significant relationship was observed in the patients with mixture of full/partial edentulism and the patients with teeth problems, possibly due to the small sample size in these groups. CONCLUSION: This study revealed that the multi-axis assessment protocol was sufficiently valid to predict the level of treatment difficulty in prosthodontic care in patients with fully edentulous defects and with partially edentulous defects.