To test the fracture toughness of alumina; a Surface-Crack-in-Flexure (SCF) method, a Single-Edge-Precracked-Beam (SEPB) method and a Single-Edge-V-Notched-Beam (SEVNB) method were used at crosshead rates ranging from 0.005 mm/min to 2 mm/ min and relative humidity ranging from $15\%;to;80\%$. The results show that the fracture toughness tested by the SCF method increases with either an increasing loading rate or decreasing relative humidity; in contrast, the toughness by the SEPB method and the SEVNB method does not depend on the loading rate or the relative humidity. Theoretical analysis of the way slow crack growth affects the apparent fracture toughness indicates that the three testing methods have different effects with respect to the loading rate and the relative humidity; moreover, these differences are attributable to differences in the size of the cracks or notches.