The objectives of the study are to demonstrate the non-inferiority of PG201 (Layla(®)) 600 mg in comparison with celecoxib 200 mg for the treatment of symptomatic knee osteoarthritis (OA). In total, 309 patients were randomly assigned to receive either the test drug, PG201 600 mg (n = 154) or celecoxib 200 mg (n = 155). The primary efficacy variable was improvement in mean 100-mm pain VAS score from baseline to the final visit (week 8), and this value was compared between the 2 treatment groups. Secondary outcome variables included changes from baseline in the Western Ontario and McMaster Universities Arthritis Index (WOMAC) pain VAS score and subscale score, patient's global assessment of disease status quality of life (short form-36) and responder index at weeks 4 and 8. For safety assessment, adverse events were recorded at each clinical visit. At weeks 8, the 100-mm pain VAS scores were significantly decreased in patients receiving both PG201 600 mg (p < 0.0001) and celecoxib 200 mg (p < 0.0001) as compared to the baseline scores; however, no statistically significant differences in these values were noted between the groups (p = 0.312). These results met pre-specified criteria for non-inferiority for both the intent-to-treat and per-protocol populations. PG201 600 mg and celecoxib 200 mg were both well tolerated and no statistically significant differences in the tolerability profile between the groups. PG201 600 mg was as effective and safe as celecoxib 200 mg in the treatment of symptomatic knee OA and might be a useful new medication for the treatment of symptomatic knee OA.