High-temperature slow-crack-growth behaviour of hot-pressed silicon carbide was determined using both constant-stress-rate ("dynamic fatigue") and constant-stress ("static fatigue") testing in flexure at 1300 degrees C in air. Slow crack growth was found to be a governing mechanism associated with failure of the material. Four estimation methods such as the individual data, the Weibull median, the arithmetic mean and the median deviation methods were used to determine the slow crack growth parameters. The four estimation methods were in good agreement for the constant-stress-rate testing with a small variation in the slow-crack-growth parameter, n, ranging from 28 to 36. By contrast, the variation in n between the four estimation methods was significant in the constant-stress testing with a somewhat wide range of n=16 to 32.