Twelve aerial counts on 1,500 square miles, including the southern two-thirds of Murchison Falls National Park, Uganda, and vicinity, January, 1957, to August, 1959, showed an average of 7,000 to 8,000 elephants (Loxodonta africana) over the 32-month period. These observations did not substantiate an earlier hypothesis that elephants migrated into the Park during the dry season in January and February and emigrated with the long rains in May and June. Although migratory movements were related to rains, they did not generally follow an annual cycle. The data show a stable population of about 4,100 elephants from January through August, 1957; significant immigration beginning with the short rains in late September until the population doubled by December, 1957; a decline by about 1,500 head during the dry season in January and February of 1958; a great increase to a peak count of 12,389 in late May and early June of 1958 (correlated with the long rains); emigration to a low of 5,461 in February, 1959; immigration to about 8,400 (again correlated with the long rains in May) and stabilization of the population at that level until termination of the investigation in August. The over-all trend spans more than a year; but in 2 calendar years immigrations were associated with long, heavy rains in May, and emigrations were significant during the long dry season in January and February. The migratory behavior of the elephants was affected by the harassment of shooting to control damage to forests and agricultural crops, from the 1930's onward. Vegetational changes on Butiaba Flats, where forage apparently was depleted for a period of time and subsequently began recovering, may also have influenced behavior. Seemingly, the elephants have been concentrating in ever-increasing numbers in the Park area, and it was evident at the termination of the study that numbers must be regulated to avoid damage to the vegetation and the future welfare of the population of elephants. As a guide to regulation of numbers, an increment of 7 to 8 percent provides a conservative basis for determining annual harvests. Three counts in which a concerted effort was made to ascertain the number of calves showed 7.1, 7.2, and 7.3 percent young, with no indication of any significant peaks of parturition during the calendar year.