Abstract: Abstract: Emitter clogging is an unavoidable problem when the hard water is used for drip irrigation and it is tightly related to the form and size of the internal structure. This study was designed to evaluate the performance of 6 types of emitters for drip irrigation by using ultrapure water and hard water, and main characteristic parameters influencing clogging were determined. A total of 6 emitters were used for the test. Among them, 3 were non-pressure-compensated, 2 were molded and welded into dripline wall, and 1 was inserted into thick wall. The other 3 emitters, being inserted into thick wall, were all pressure-compensated. The hard water was prepared at 500 mg/L using ultrapure water and chemical reagents (CaCl2 and NaHCO3). The experiment was carried out at the Irrigation Hydraulics Laboratory of Northwest A&F University, from October 2014 to January 2015 and from May 2015 to July 2015. A drip irrigation system with cyclic water was built, which had 12 laterals. Each of these emitters was installed to the 2 of laterals. The drip irrigation system was operated at a regime of 4 h operation a week under a regulated pressure of 100 kPa. The water delivered for each emitter was collected for 10 min to measure the flow rate of the emitters every other week. The emitter's average relative flow rate and the emitter clogging ratio were calculated for 26 weeks to evaluate the emitter performance. The ordinal regression analysis was used to analyze the influence of characteristic parameters on emitter clogging. The emitter characteristics used included pressure-compensation, nominal flow rate, water passage length, water inlet fence and the minimum size in the width and depth of flow passage section. All of the emitters were dissected after all the irrigation events. Photos of the emitter interior structures were taken by a camera (Canon PowerShot SX500 IS). Then the clogging was assessed by information of sediment deposition inside the emitter. The results showed that the decline in average relative flow rate was observed with 6 kinds of emitters during the drip irrigation system running time. Emitter E1 and E6 had the best performance. On the contrary, emitter E4 showed the worst performance. And emitter E3 was observed a highest emitter clogging ratio after 2 weeks. Ordinal regression analysis showed that the effect of the minimum size of flow passage section on the emitter anti-clogging ability was statistically significant (P<0.01). And the anti-clogging ability of emitter had a positive correlation with the minimal size of flow passage section. As the minimal size of flow passage section increased, the emitter anti-clogging ability was improved. The other emitter characteristics were not found to be statistically significant (P>0.05). Dissection of emitters revealed that clogged occurred at different areas for different emitters. E1 was mainly clogged at the head of the labyrinth flow passage. E2 was clogged at the fence where the water entered from. E3 was mainly clogged at the head of the labyrinth flow passage behind the fence. E4, E5 and E6 were mainly clogged at a little narrow groove, which was used to achieve a function of pressure compensation. The emitter clogging of 6 types mainly occurred at the flow passage section with the minimal size. When there was more than 1 area with the minimal size of flow passage section inside the emitter, clogging had tended to occur at an area that particles was easily to deposit on. This area was called by "low velocity area". Thus, choosing emitters with the larger water passage section and avoiding the flow passage easily causing deposition are important when hard water is used for drip irrigation. These results can provide important information for understanding emitter clogging mechanism and some suggestions for scientific emitter selection of hard water irrigation system.