Objective: Recently we reported that glucosamine sulphate (GS) did not have an effect on the symptoms and progression of primary care patients with hip osteoarthritis (OA). The aim of this present study was to investigate whether there are subgroups of patients with hip OA for whom GS might be an effective therapy. Method. We randomized 222 patients with hip OA that met one of the American College of Rheumatology criteria to either 1500 mg of oral GS or placebo once daily for 2 years. Subgroup analyses were predefined for radiographic severity (Kellgren & Lawrence (KL) = 1 vs >= 2) and for type of OA (localised vs generalised). Additional exploratory subgroup analyses focused on groups based on pain level, pain medication use, baseline joint space width (JSW), and concomitant knee CA at baseline. Primary outcome measures were Western Ontario MacMaster Universities (WOMAC) pain and function scores over 24 months, and joint space narrowing (JSN) after 24 months. Results: In the predefined subgroups based on radiographic severity and type of OA, the outcomes WOMAC pain, function and JSN were similar for the GS and placebo group. Conclusion: GS was not significantly better than placebo in reducing symptoms and progression of hip OA in subgroups of patients. (C) 2008 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.