Historical records of freeze-up and break-up and aerial observations of these events have been collected for the Mackenzie River between Great Slave Lake and the Beaufort Sea. These data have been studied with reference to the progress of ice formation and disintegration, the effects of break-up on flooding in the Mackenzie Delta, and the influence of flow travel time on freeze-up and break-up dates. The progress of break-up and freeze-up along the Mackenzie has been analyzed both by standard statistical procedures (Wilcoxon matched-pairs signed-ranks test, and Kendall coefficient of concordance) as well as by aerial mapping (1959, 1961, and 1962). Results show that the Mackenzie first opens up at its outlet from Great Slave Lake and also where major tributaries enter it. Break-up is nearly simultaneous between Fort Simpson and Fort Good Hope or even Arctic Red River, but there is normally a delay of a few days between Fort Providence and Fort Simpson. The Mackenzie Delta breakup is complex because the southwestern partis controlled by the opening up of the Peel River and mountain rivers. Freeze-up is more orderly than break-up as it progresses fairly regularly upstream from the mou th to Great Slave Lake. The variability of channel break-up and numerous ice jams caused severe flooding in the Mackenzie Delta during the latter part of May and early June 1961. Visual estimates suggest that over 95 per cent of the delta was under water. The estimated flow travel time between Great Slave Lake and the Beaufort Sea is 15 to 20 days during the open season. The progress of break-up cannot be directly related to the normal flow travel time due to the changes in stage that occur during the break-up period. Water temperatures· and flow travel time do not provide a satisfactory explanation for the progress of freeze-up, and further study on the heat budget of the Mackenzie River is necessary.