Transient Reactor Test Facility (TREAT) Test H6 was run to simulate a transient overpower (TOP) initiated 50 cent/s hypothetical core disruptive accident (CDA). The primary purpose was to investigate the extent to which molten fuel could be removed from the active core region following fuel pin failure, and the extent to which this would be accomplished while maintaining coolant flow. The hydraulic system of the Mk-IIC integral loop used for the H6 test was such that coolant flow rates and pressures typical of those in the Fast Flux Test Facility would be attained. The test fuel sample consisted of a bundle of seven mixed-oxide fuel pins which had been preirradiated in the Experimental Breeder Reactor II to ∼6 at.% burnup. The liquid sodium coolant had an initial velocity of 6.20 m/s at a temperature of 742 K. A programmed TREAT power ramp with a period of 1.65 s was used to bring the experimental fuel sample to failure conditions. The test data showed that there were three main events associated with fuel pin failure. During the first of these events, fuel was removed from the active fuel region and relocated ∼40 cm downstream. The coolant flow rate recovered to ∼93% of its preevent value. Additional fuel was removed from the active fuel region during the second event and again relocated some 40 cm downstream. However, molten fuel also began to accumulate in a region centered on the centerline of the original fuel column. The coolant flow rate recovered to ∼75% of its initial value. The third event was considerably more violent than the others and while a considerable quantity of fuel was relocated well downstream of the active fuel column, a blockage was formed at the top of the fuel column which reduced the coolant flow to zero. The test was terminated at this time. Analysis showed that the first fuel pin failure occurred when the areal fraction of fuel above the solidus was ∼0.5, and the fuel pin cladding temperature was ∼950 K. From examination of thermocouple data, in conjunction with thermal-hydraulic analysis, it appeared that the location of the first two events was at the fuel axial midplane, while the location of the third event was probably close to the top of the fuel column. Finally, analysis of the flowmeter signals indicated that the fuel pin holder failed during the third event. This could be at least partially responsible for the coolant channel blockage following this event. Through the first two failure events, however, the H6 test demonstrated, for the first time where preirradiated fuel was being used, that fuel could be removed from the active core region while general coolability was maintained.