Abstract: In order to reduce the misjudgment of results out of specification and the frequency of detection, the indirect method of purge and trap-gas chromatography (P&T-GC) and the direct method of liquid-liquid extraction-gas chromatography (LLE-GC) were used for determination of 35 finished water samples and pipe network water samples in 7 cities with their determined results of chloral greater than limit (10 μg·L^-1) specified by national standard of GB 5749-2006, and the reasons for the difference between the results of these two methods were also analyzed. Results of method verification showed that both of two methods could meet the quality control requirements, but the parallelism of P&T-GC was better than that of LLE-GC. Determined results of these two methods showed that results of chloral in 57.1% of samples determined by P&T-GC were over 10 μg·L^-1, while results obtained by LLE-GC were less than the specified limit. The t-test results showed that a significant difference was found between the two methods, and the reasons for it were analyzed. As shown by results that the stronger the ability of trichloroacetic acid coexisting in water to be converted into trichloromethane by derivatization, the higher the content of chloral determined by P&T-GC, but this trend was not found in LLE-GC; the farther the distance between the sampling point and the outlet of the water plant was, the higher the content of chloral determined by P&T-GC and LLE-GC, but the increase of P&T-GC was about 5 times than that of the latter. Therefore, LLE-GC was recommended to retest and confirm results out of specification of actual water sample determined by P&T-GC.