We investigated the characteristics of patients who achieved Japanese-style deep flexion (seiza-sitting) after total knee replacement (TKR) and measured three-dimensional positioning and the contact positions of the femoral and tibial components. Seiza-sitting was achieved after surgery by 23 patients (29 knees) of a series of 463 TKRs in 341 patients. Pre-operatively most of these patients were capable of seiza-sitting, had a lower body mass index and a favourable attitude towards the Japanese lifestyle (27 of 29 knees). According to two-/three-dimensional image registration analysis in the seiza-sitting position, flexion, varus and internal rotation angles of the tibial component relative to the femoral component had means of 148° (. sd. 8.0), 1.9° (. sd. 3.2) and 13.4° (. sd. 5.9), respectively. Femoral surface contact positions tended to be close to the posterior edge of the tibial polyethylene insert, particularly in the lateral compartment, but only 8.3% (two of 24) of knees showed femoral subluxation over the posterior edge. The mean contact positions of the femoral cam on the tibial post were located 7.8 mm (. sd. 1.5) proximal to the lowest point of the polyethylene surface and 5.5 mm (. sd. 0.9) medial to the centre of the post, indicating that the post-cam contact position translated medially during seiza-sitting, but not proximally. Collectively, the seiza-sitting position seems safe against component dislocation, but the risks of posterior edge loading and breakage of the tibial polyethylene post remain. Cite this article: Bone Joint J 2013;95-B:782–7