OBJECTIVES: To determine the financial impact of inclusion of glucosamine in the hospital formulary of Petchabun Hospital. METHODS: Conventional NSAIDs, celecoxib, glucosamine were included in the analysis. Hospital perspective was used and we considered only patients who failed to control pain by acetaminophen. Treatment cost of GI and CV event was calculated based on decision tree model in which cooperated with probability from the literature and local DRG cost data. Average cost of drug use per patient per year was calculated from the medical history and computerized dispensing data. Delphi technique with all treating orthopaedics was used to obtain the estimates of number of patient eligible for glucosamine use and effect on volume of other drugs' use in the following years. Probabilistic analysis was used to capture the certainties around estimations. We analyzed this data in a 5-year timeframe (2005-2009) as we assumed the steady penetration of glucosamine was reached.  RESULTS: In 2005, we estimated there was an increase in drug budget for knee osteoarthritis treatment around 0.8 million baht (23,000 US$) which was growing from 2004 at 7%; it was similar to the growth rate in the ealier year. This was mainly due to the expected reduction in NSAIDs and coxib uses. The forecast budget impact in 2006-2009 were about 0.8, 0.6, 0.5, and 0.2 million baht at the growth rate 6.2%, 4.4%, 3.5% , and 1.5%respectively.  Cost of glucosamine was found to be the most sensitive variable, followed by cost of celecoxib and number of patients using glucosamine. CONCLUSIONS: Hospital administrators found that glucosamine use resulted in an affordable financial burden to the drug budget for knee osteoarthritis in the hospital. However, to ensure the effective use of glucosamine, they developed the guideline for glucosamine use and also monitor the clinical and economic outcomes. Updated analyses were also recommended to obtain the reliable information for budget planning in the following years.