There is no consensus on the most reliable method of ascertaining falls among the elderly. Therefore, we investigated which method captured the most falls among prefrail and frail seniors from 2 randomized controlled trials conducted in Zurich, Switzerland: an 18-month trial (2009-2010) including 200 community-dwelling prefrail seniors with a prior fall and a 12-month trial (2005-2008) including 173 frail seniors with acute hip fracture. Both trials included the same methods of fall ascertainment: monthly active asking, daily self-report diary entries, and a call-in hotline. We compared numbers of falls reported and estimated overall and positive percent agreement between methods. Prefrail seniors reported 499 falls (fall rate = 2.5/year) and frail seniors reported 205 falls (fall rate = 1.4/year). Most falls (81% of falls in prefrail seniors and 78% in frail seniors) were reported via active asking. Among prefrail seniors, diaries captured an additional 19% of falls, while the hotline added none. Among frail seniors, the hotline added 16% of falls, while diaries added 6%. The positive percent agreement between active asking and diary entries was 100% among prefrail seniors and 88% among frail seniors. While monthly active asking captures most falls in both groups, this method alone missed 19% of falls in prefrail seniors and 22% in frail seniors. Thus, a combination of active asking and diaries for prefrail seniors and a combination of active asking and a hotline for frail seniors is warranted.