Abstract This study was designed to evaluate the effect of silage source (barley vs. wheat silage) when harvested at two chop lengths (low vs. high physically effective neutral detergent fiber [peNDF]) and when barley silage was partially replaced with straw to increase the undigested neutral detergent fiber (uNDF) concentration on performance and carcass characteristics of finishing steers. Four hundred and fifty yearling commercial crossbred steers with an initial body weight (BW) of 432 ± 30.5 kg were allocated to 30 pens and fed diets containing 90% concentrate:10% forage for 123 d in a completely randomized block design with a 2 × 2 + 1 factorial arrangement. Treatments included 1) barley silage (BarS) with low peNDF (LpeNDF); 2) BarS with high peNDF (HpeNDF); 3) BarS with straw to yield a diet with LpeNDF + uNDF; 4) wheat silage (WhS) LpeNDF; and 5) WhS HpeNDF. There were no silage × peNDF interactions for dry matter intake (DMI), average daily gain (ADG), or gain to feed ratio (G:F), but cattle fed WhS LpeNDF had a lower (P < 0.01) proportion of yield grade 3 and a greater proportion in yield grade 2 carcasses than cattle fed BarS LpeNDF or HpeNDF and WhS HpeNDF. Cattle fed WhS LpeNDF had greater (P = 0.02) incidence of severe liver abscesses when compared with cattle fed BarS LpeNDF or HpeNDF and WhS HpeNDF. Cattle fed BarS consumed less (P < 0.01) uNDF as a percentage of BW, had increased (P = 0.02) ADG, heavier (P = 0.02) hot carcass weight, with greater (P = 0.01) back fat thickness, and (P < 0.01) incidence of minor liver abscesses when compared with cattle fed WhS. Feeding HpeNDF did not affect DMI, ADG, or G:F, but increased (P = 0.02) marbling score and reduced (P < 0.01) the proportion AA quality grade and increased (P < 0.01) those classified as AAA when compared with cattle fed LpeNDF. Cattle fed low uNDF had lesser (P < 0.01) uNDF intake as a percentage of BW, greater dressing percentage (P = 0.01), had a lower (P < 0.01) proportion of carcasses in yield grade 2, and a greater (P < 0.01) proportion of carcasses in yield grade 3 when compared with cattle fed high uNDF. Thus, silage source, peNDF, and uNDF content do not impact DMI or G:F when diets contain 10% forage, but BarS relative to WhS as well strategies increasing the peNDF concentration may increase ADG, HCW, back fat thickness, dressing percentage, marbling score, and carcasses classified as quality grade AAA. Future research is needed to evaluate the usefulness of peNDF and uNDF in rations for finishing cattle.