Formerly sedentary older women (M age = 62 yr.) were recruited for either a wait-list control group (n = 40) or a 10-.wk., 3 days/wk. weight-training treatment group (n = 48). Only the treatment group demonstrated significant before- to after-exercise improvements on the Exercise-induced Feeling Inventory scales of Positive Engagement, Revitalization, Tranquility, and Physical Exhaustion (ds = |.43| to |.96|). For the treatment group, scores on Physical Self-concept of the Tennessee Self-concept Scale: 2 were significantly correlated (r = -.28) with after-exercise changes in Physical Exhaustion. Muscular strength was significantly correlated with after-exercise changes in scores on three Exercise-induced Feeling Inventory scales (rs = |.28| to |.31|). Simultaneous entry of Physical Self-concept scores and muscular strength into multiple regression equations increased the explained variance in Exercise-induced Feeling Inventory score changes for each of its four scales. Statistical significance was only reached, however, on changes in the Physical Exhaustion scale (R2 = .16, p = .02). Limitations and implications for theoretical development and applied use were discussed.