Wear of the polyethylene tibial insert remains a concern in total knee replacement (TKR) because the debris generated can lead to osteolysis and implant loosening. The purpose of this study was to determine if linear penetration is an accurate surrogate for volume loss due to creep and wear in TKR. The study was conducted on a group of 83 retrieved tibial inserts of a contemporary cruciate-retaining TKR design. Insert volume loss was determined with a recently developed and validated method combining metrology with an autonomous mathematical reconstruction algorithm. Penetration was determined in two ways: using an autonomous mathematical reconstruction (P_AMR) and from minimum insert thickness measurements using a dial indicator (P_dial). Volume loss correlated strongly to P_AMR (R² = 0.39 − 0.57, p<0.0001) and P_dial (R² = 0.22 − 0.46, p<0.0001), the correlation markedly improving with the removal of just two or three outliers. However, penetration was only a fair predictor of volume loss for individual tibial inserts, yielding 95 % prediction interval bands of ± 42.5 mm³ to ± 55 mm³. Multiplying penetration by a measure of wear scar surface area (e.g., wear scar medial-lateral width and anterior-posterior length) led to a better predictor variable for volume loss. More complex regression models that take into account the location of the penetration maximum and insert size improved volume loss prediction but only marginally if outliers, such as those stemming from edge locations, are avoided. The obtained prediction intervals apply only to the knee design evaluated here.