Objective: Previous research using personality traits to predict life outcomes has typically utilized the Big Five factors and, occasionally, their facets. However, recent research suggests that using items (reflecting personality nuances) can account for greater predictive variance. The present study examines the predictive validity of the different levels of the personality trait hierarchy (factor, facet, and nuances). Method: Confirmatory Factor Analyses (CFA) were performed on the data (N = 440) to confirm the structures of the Big Five levels prior to using Elastic Net Regression (ENR; with 10-fold cross-validation and shrinkage parameter) to predict outcomes at the factor, facet, and item level. Models were trained and applied for prediction in separate samples. Results: The results showed that nuances, on average, provided greater explained variance (34%) than both facets (22.5%) and factors (12%) for all six outcome predictions, suggesting that narrower traits are more effective in predicting outcomes than the Big Five factors. Conclusion: Findings suggest that there may be benefits to using narrower characteristics for predicting outcomes when predictive validity is the goal. Implications, limitations, and directions for future research are discussed.