INTRODUCTION: Subsidence has been noted with both cemented and uncemented stems in hip arthroplasty. On most occasions, it's minimal (i.e. less than 2 mm) and stabilises at 1 year. However, when its progressive and significant, it causes loss of length and horizontal offset, and when symptomatic warrants a revision. MATERIALS AND METHODS: A retrospective radiological review of the patient database was carried out to identify all patients with age ≥ 70 years who underwent elective hip arthroplasty using uncemented HA coated collared stem and had radiographs at 1 year follow up. A total of 176 patients were identified from January 2009 to June 2010. 2 independent investigators classified the proximal femur based on Dorr type and calculated the subsidence based on Engh and Massin method of calculating the distance between the tip of greater trochanter and shoulder of the prosthesis. RESULTS: 7 patients (4 Bs, 2 As, 1 C) had a subsidence of ≥ 2 mm (2-3.2 mm) at 1 year. None of them was symptomatic. 1 of them was secondary to a missed calcar crack and continued to subside for 9 mm before it stabilised on the lesser trochanter at 4 years follow up. There were Dorr 22 (12.5%) type A, 147(83.5%) type B and 7(4%) type C. The mean age was 77.4 years (70-91 years) and male: female ratio was 7:15. 3 patients had an intraoperative calcar crack requiring cabling. All were mobilised full weight-bearing postoperatively, and none had a subsidence of >2 mm at 1 year follow-up. DISCUSSION: Our subgroup analysis identified that subsidence can happen when the collar is "non-functional" and the initial press fit of the stem wasn't achieved. It can also occur in an event of calcar crack, which is missed intraoperatively. In cases of calcar crack which went on to have to cable during the primary procedure, it neither changed the post-operative rehabilitation nor did it show an increased risk of subsidence. CONCLUSION: A fully hydroxyapatite (HA) coated collared stem, when used in elderly age group for elective THR, has only 2% risk of intraoperative periprosthetic fracture. There's a 4% risk of radiologically significant subsidence (i.e. ≥2 mm), however, it has not proven to be clinically significant in our study. Dorr canal type had no bearing on either risk of periprosthetic fracture or subsidence. Collared stems did not have a statistically significant difference in risk of subsidence and peri-prosthetic fracture in comparison to un-collared stem, although there was a non-significant trend in favour of collar use.