Our objective was to investigate the association between feed bunk refusals (FBR) with feed conversion efficiency (FCE) in 2 commercial dairy farms with compost bedded pack barns in Argentina (Chiavassa Dairy, CD; LaCenobia Dairy, LD). For 5 mo (September 2017 to January 2018), we daily registered DM offered (DMO, kg/d), DMI (kg/d), milk yield (MY, kg/d), FCE (kg milk/kg DMI), and FBR (%DMO) in 4 pens (0 to 150 DIM; 230 and 150 cows/pen at CD and LD, respectively). Milk yield was averaged by pen and DMI was estimated by difference between feed offered and refused. TMR DM content was determined with a forced-air oven for 2 h at 135°C. At CD, cows were milked 3x/d and fed 2x/d with a TMR composed by corn silage, alfalfa silage, alfalfa hay, corn grain, soybean meal, and mineral premix that averaged 49.8% DM, 15.8% CP, 29.9% NDF, and 2.89 Mcal ME/kg DM. Milk yield, DMI, and FBR (mean ± SE) was 42.06 ± 6.01 kg/d, 23.88 ± 2.42 kg/d, and 6.90 ± 4.38%DMO, respectively. At LD, cows were milked 2x/d and fed 1x/d with a TMR composed by corn silage, alfalfa silage, wheat silage, corn grain, soybean meal, peanut meal, and mineral premix that averaged 46.0% DM, 16.3% CP, 31.9% NDF, and 2.41 Mcal ME/kg DM. Milk yield, DMI, and FBR (mean ± SE) was 26.22 ± 2.15 kg/d, 22.55 ± 2.31 kg/d, and 6.45 ± 5.93%DMO, respectively. We run Pearson correlations to evaluate association between FCE and MY, DMI, and FBR (InfoStat, 2017). Effect of FBR level (0–3%, 3–6%, and >6%) was also evaluated with a mixed linear model with FBR level and day as fixed and pen within dairy as random effects. For both dairies, FCE was positively (P < 0.05) correlated with MY (r = 0.50 CD; r = 0.24 LD) and FBR (r = 0.38 CD; r = 0.74 LD) and negatively (P < 0.05) correlated with DMI (r = - 0.33 CD, r = - 0.68 LD). FBR level affected (P < 0.05) DMI and FCE, but it did not MY (P > 0.05). With FBR >6% (vs. 3 – 6% and 0 – 3%), FCE increased (1.53 vs. 1.43 and 1.39 ± 0.29 kg milk/kg DMI) and DMI reduced (22.43 vs. 23.65 and 24.43 ± 1.14 kg/d). Our data shows the importance of FBR management to maximize FCE in commercial dairy farms. More appropriate FCE should be estimated using energy corrected milk.