OBJECTIVES: To document the psychometric properties and estimate a minimal important difference (MID) for the Sinonasal Outcome Test-16 (SNOT-16) in patients with acute bacterial sinusitis. METHODS: Secondary psychometric analysis was undertaken on data from a phase III trial to evaluate the efficacy and safety of moxifloxacin over 5 consecutive days in the treatment of acute bacterial sinusitis (N=374). Trial patients also completed the RAND SF-36, the Activity Impairment Assessment (AIA) and a global rating of change.  The SNOT-16 has 16 items and is scored as a uni-dimensional measure of symptoms of rhinosinusitis.  Internal consistency and construct validity (against SF-36 and AIA) were assessed at baseline.  Sensitivity and MID were assessed by comparing test of cure data against baseline using statistical methods (effect size (ES), standardised response mean (SRM); and standard error of measurement (SEM) and ½ standard deviation (½SD)) and anchor methods against the smallest change on the global rating of change.  Missing data were imputed using last observation carried forward. RESULTS: Internal consistency (α=0.874) and construct validity (SF-36 Role Function r=-0.44; Vitality r=0.45, Social Function r=0.48, Pain r=0.51, AIA r=0.67 all P<0.001) were confirmed.  The SNOT-16 was sensitive (ES =1.68; SRM=1.36).  The MID estimated by the anchor method was -0.85; but MID was lower by distributional estimates (SEM=0.21, ½SD = 0.29).  CONCLUSIONS:The study confirms the good psychometric properties of the SNOT-16 in patients with acute bacterial sinusitis and demonstrates the sensitivity and MID for this measure. The 3 MID estimates were widely separated and the higher anchor based estimate was settled on.