The following study was designed to find out the association of the mid-upper arm body composition and body mass index (BMI) with depressive symptoms among the old age home based (group 1) and family based elderly (group 2) subjects. In group 1 elderly males, 51.5% were showing mild depressive symptoms and 17% were showing moderate to severe depressive symptoms. On the other hand, among group 2 elderly males, prevalence of mild depressive symptoms was in 27% and moderate to severe depressive symptoms was in 4.5% of the subjects. Among elderly with and without depressive symptoms, BMI, waist-hip ratio (WHR), arm muscle circumference (AMC), arm muscle area (AMA), arm fat area (AFA), fat mass (FM), muscle mass and bone mass were significantly (p < 0.01) lesser among group 1 as compared to group 2. Among group 1, the AFA (R2 change = 0.178), AMC (R2 change = 0.035), AMA (R2 change = 0.032), FM (R2 change = 0.022) and muscle mass (R2 change = 0.019) have highest contribution to the variance of depressive symptoms. On the other hand, BMI (R2 change = 0.04) and WHR (R2 change = 0.03) had the highest, though small, contribution to the variance of depressive symptoms among group 2. The upper arm body composition indices were better indicator of depressive symptoms in group 1, and BMI and WHR were better indicators of depressive symptoms in group 2. Therefore, among group 1 subjects, the upper arm body composition indices prove to be better regular health monitoring indices as compared to traditionally used BMI and WHR.