<p>Reliable interpretation of bored pile capacity from full-scale load-test data is essential for safe, economical design, yet evaluations across sand, clay and layered soils remain limited. A database comprising 75 full-scale axial compression tests on bored concrete piles which evaluated twelve interpretation criteria. All tests reached plunging failure or a large-displacement response, defined by pile-head settlement ≥ 10% of pile diameter where no clear plunge occurred. Method performance was assessed using applicability, pile geometry effects, failure settlement and statistical metrics. Measured-to-interpreted capacity ratios showed larger apparent slope variation with pile diameter than length; however low R² values ≤ 0.016 for diameter and ≤ 0.106 for length, indicated weak size-related trends. Soil-specific comparisons showed that DeBeer performs best in clay and layered soils, whereas Mazurkiewicz performs best in sand and all soils combined considering statistical performance and applicability. DeBeer and Mazurkiewicz gave the closest agreement with measured capacities, with high applicability, median R²≈ 0.99, near unity slopes and low errors; therefore, their arithmetic average is proposed as a unified criterion. This averaging is physically motivated because DeBeer captures early yielding, while Mazurkiewicz reflects later failure response, reducing opposite bias. Weighted-average sensitivity analysis confirmed that the unweighted DeBeer-Mazurkiewicz average gave the lowest RMSE and reduced the mean prediction ratio from 0.967 for DeBeer and 1.050 for Mazurkiewicz to 1.009 for the proposed average. Targeted partial-curve validation indicated potential applicability to non-failed load-settlement curves. The proposed average criterion offers a practical unified approach for interpreting bored pile load tests across sand, clay and layered profiles.</p>