Purpose: This paper aims to evaluate the predictive validity of the Competency Coefficient (K), a behavioral indicator derived from structured assessments of automotive R&D project managers, by examining its correlation with objective project performance outcomes. Methodology: The study performs a statistical analysis of K against five z-standardized Key Performance Indicators (KPIs) to identify the relationship between behavioral competencies and project performance. Predictive validity was evaluated using Pearson/Spearman correlations and Ordinary Least Squares (OLS) regression; robustness in small samples and model adequacy were assessed with 10,000-sample bootstrap intervals, Leave-One-Out Cross-Validation (LOOCV),Prediction Sum of Squares (PRESS) (PRESS/Q²), and tests for quadratic nonlinearity. Findings: Results reveal positive and statistically significant associations between the Competency Coefficient (K) and KPI-based performance indices. The linear model explained roughly 93% of the variance in project results, and cross-validation confirmed consistent out-of-sample performance (Q² = 0.88). The restricted sample size (n = 7) and singular organizational environment limit generalizability, while contextual factors may also influence the reported outcomes. Originality/Value: The paper provides original empirical evidence that competency-based behavioral indicators can function as dependable, measurable elements of project performance assessment. The findings emphasize methodological feasibility rather than universal applicability. The contribution lies in the measurement and validation technique, which may be duplicated for verification in larger and more diverse samples.