Objective: Our objective was to determine the effects of initiating narasin, at different production stages, on growing-finishing pig performance. Materials and Methods: In Exp. 1, 360 barrows (BW = 23.0 +/- 0.13 kg) were used in a completely randomized design (45 pens; 15 replications) with 90-d treatments: (1) an unmedicated diet (CON), (2) 15 mg/kg narasin (NAR-90d), or (3) 11 mg/kg virginiamycin (VIR-90d). In Exp. 2, 480 barrows (BW = 26.3 +/- 0.14 kg) were used in a completely randomized design (60 pens; 15 replications) with the following treatments: (1) unmedicated CON diet, (2) 15 mg/kg narasin (0 to 91 d; NAR-91d), (3) CON (0 to 56 d) followed by 15 mg/kg narasin fed from 56 to 91 d (NAR-35d), or (4) 27.6 mg/kg zinc bacitracin (0 to 91 d; ZB-91d). In Exp. 3, 416 barrows and gilts (BW = 84.3 +/- 0.39 kg) were used in a randomized complete block design (52 pens; 26 replications) with 35-d treatments: (1) unmedicated CON diet or (2) NAR-35d. Results and Discussion: In Exp. 1, pigs fed NAR-90d had greater BW (d 90) and ADG (d 0 to 90) compared with pigs fed the CON dietary treatment. The ADG of pigs fed NAR-90d were greater than those of pigs fed the CON treatment on d 0 to 28 and d 28 to 56 and VIR-90 on d 0 to 28. Pigs fed NAR-90d had greater hot carcass weights and yields than those fed CON or VIR-90d. In Exp. 2, pigs fed NAR-91d (d 0 to 56) and pigs fed NAR35d (d 56 to 91) had greater ADG and ADFI than any of the other treatments. In Exp. 3, pigs fed NAR-35d had greater G:F than pigs fed the CON treatment. Implications and Applications: Pigs fed narasin tended to have greater growth performance responses in the periods after first receiving narasin. However, when narasin is fed for the entire growing-finishing period, pigs may have carcass responses in some cases. Improvements in growth performance were realized whether narasin was initiated in the early grower or finishing phase.