Few studies have evaluated the benefit of providing exercise to underprivileged older adults at risk for falls. Economically and educationally disadvantaged older adults with previous falls (mean age 79.06, SD= 4.55) were randomized to four months of multi-modal exercise provided as fully supervised center-based (FS, n =45) or as minim ally supervised home-based (MS, n = 42) or to non-exercise controls (C, n= 32). Comparing groups on the mean change in fall-relevant mobility task performance between baseline and four months and compared to the change in C, both FS and MS had significantly greater reduction in Timed Up and Go F(2,73)= 5.82, p =.004, η2partial = 0.14 and increase in Tandem Walk Speed F (2,73)= 7.71, p <.001 η2partial = 0.17. Change in performance did not statistically differ between FS and MS. In community dwelling economically and educationally disadvantaged older adults with a history of falls, minimally supervised home-based and fully supervised center-based exercise programs may be equally effective in improving fall-relevant functional mobility.