The elevated metabolic cost of walking (C_W) in elderly adults has been hypothesised to be a consequence of the declines in muscle mass, strength and stability (Malatesta et al, 2003). Exercise training partially reverses these declines; does it influence C_W? Community dwelling elderly adults were assigned to a training (TRA, mean ± SD age = 73 ± 3, n = 22) or a non-training control group (CON, age = 73 ± 4, n = 12). The intervention lasted 12 months. TRA performed 1 hour supervised exercise sessions twice per week. Leg press, leg extension, calf raise, chest press, and seated row exercises were performed on resistance machines (2-3 sets of 8-10 repetitions at 80-100% of 8 repetition maximum) in addition to aerobic, balance, and flexibility exercises. In addition to supervised sessions, home based sessions utilising resistance bands and light aerobic exercises were performed once per week. Before and after the intervention, to assess C_W, the net (gross − standing) oxygen required to walk unit distance was measured using an automated analyser (Cosmed K4b²) whilst participants walked on a motor driven treadmill at 0.83, 1.11 and 1.39 m/s. Isometric knee extensor muscle strength (torque at 90 deg), single leg balance time, and 6 min walk distance were also measured to assess changes in traditional measures of functional capacity. Interactions between group and time were assessed using ANOVA. The study had local ethical committee approval. At baseline, mean C_W of the elderly adults in this study was 34% higher than in a convenience sample of young adults (age = 27 ± 3, n = 23). However, the intervention had no effect on C_W in either TRA or CON. This was despite improvements in mean knee extensor muscle strength, balance time, and 6 min walk distance in TRA (+20, +12 and +6%, respectively) that were greater than changes in CON (+5, -28 and 0%) (F_group*time, P < 0.05). In conclusion, training-induced improvements in muscle strength and balance performance were not accompanied by improvements in walking economy. These findings do not support the hypothesis that the decline in muscle mass and strength in healthy elderly adults are responsible for the elevation in C_W. Task specific training may be required to reduce the elevated metabolic cost of movement.