Model Uncertainties in Foundation Design is unique in the compilation of the largest and the most diverse load test databases to date, covering many foundation types (shallow foundations, spudcans, driven piles, drilled shafts, rock sockets and helical piles) and a wide range of ground conditions (soil to soft rock). All databases with names prefixed by NUS are available upon request. This presents the most comprehensive evaluation of the model factor mean (bias) and coefficient of variation (COV) for ultimate and serviceability limit state based on these databases. They can be used directly for AASHTO LRFD calibration. The authors provide the most complete literature survey of performance databases for other geo-structures and their model factor statistics. A practical three-tier scheme for classifying model uncertainty according to the model factor mean and COV is proposed. This empirically grounded scheme can underpin the calibration of resistance factors as a function of the degree of understanding – a concept already adopted in the Canadian Highway Bridge Design Code and the new draft for Eurocode 7 Part 1 (EN 1997-1:202x). The helical pile research in Chapter 7 was recognised by the 2020 ASCE Norman Medal.