We compared the analgesic efficacy and safety of glucosamine sulfate (GS) and chondroitin sulfate (CS) capsules or sachet preparations with glucosamine hydrochloride (GH) and CS capsules in knee osteoarthritis (OA) patients. 1,120 subjects with radiographic knee OA (Kellgren/Lawrence 2-3) were randomized (1:1:1) at 16 centers to receive GS 500 mg/CS 400 mg three times daily capsules (GI) or once daily sachet (GII) or GH 500 mg/CS 400 mg three times daily (GIII) for a 16-week trial. Primary outcome, intention-to-treat (ITT) was change from baseline of patient reported pain intensity (0-100 mm visual analogue scale) in the affected knee and variation of Lequesne's index (LI). Monthly secondary outcomes were changes from baseline in patient reported pain and LI, patient and physician global assessments of disease activity, acetaminophen consumption, and adherence. ITT population comprised 302, 301, and 306 patients in GI, GII, and GIII. Pain significantly decreased (GI = -30.9 ± 1.5; GII = -28.7 ± 1.5; GIII = -29.7 ± 1.5 mm) in all groups (P < 0.001) as well as LI (GI = -3.8 ± 0.2; GII = -3.7 ± 0.2; GIII = -3.9 ± 0.2; P < 0.001). All secondary outcomes improved (P < 0.005) for all groups. Patients that did not complete the study were 77 (44.8 %) for lack of adherence, 16 (9.3 %) consent withdrawal, 11 (6.4 %) adverse events, eight (4.7 %) lost to follow-up, and 17 (9.9 %) for other causes. Non-inferiority analysis found no differences among groups. This is a large study showing that GS/CS and GH/CS provide clinically meaningful and sustained analgesia in knee OA regardless of dose fractionation and capsule or sachet formulations.