A double-blind therapeutic investigation was performed on 178 Chinese patients suffering from osteoarthritis of the knee randomized into two groups, one treated for 4 weeks with glucosamine sulfate (GS, CAS 29031-19-4, Viartril-S) at the daily dose of 1,500 mg and the other with ibuprofen (IBU, CAS 15687-27-1) at the daily dose of 1,200 mg. Knee pain at rest, at movement and at pressure, knee swelling, improvement and therapeutic utility as well as adverse events and drop-outs were recorded after 2 and 4 weeks of treatment. The variables were recorded also after 2 weeks of treatment discontinuation in order to appreciate the remnant therapeutic effect. Both GS and IBU significantly reduced the symptoms of osteoarthritis with the trend of GS to be more effective. After 2 weeks of drug discontinuation there was a remnant therapeutic effect in both groups, with the trend to be more pronounced in the GS group. GS was significantly better tolerated than IBU, as shown by the adverse drug reactions (6% in the patients of the GS group and 16% in the IBU group--p = 0.02) and by the drug-related drop-outs (0% of the patients in the GS group and 10% in the IBU group--p = 0.0017). The better tolerability of GS is explained by its mode of action, because GS specifically curbs the pathogenic mechanisms of osteoarthritis and does not inhibit the cyclo-oxygenases as the non-steroidal anti-inflammatory drugs (NSAIDs) do, with the consequent anti-inflammatory analgesic activities but also with the several adverse reactions due to this not targeted effect. The present study confirms that GS is a selective drug for osteoarthritis, as effective on the symptoms of the disease as NSAIDs but significantly better tolerated. For these properties GS seems particularly indicated in the long-term treatments needed in osteoarthritis.