Three algorithms of multiple linear regression (MLR), k-nearest neighbor (KNN), and artificial neural network model (ANN) were used to establish a predictive model for the effect of nutrient intake in each period of Hanwoo steers on carcass traits. Also, this model pathway backward was performed to derive the appropriate nutrient intake used in Hanwoo steers. The first experiment focused on the construction of the rumen volatile fatty acid and methane prediction model. Overall, the predictive accuracy of the volatile fatty acid and methane prediction model were accurate when the KNN algorithm was used. The volatile fatty acid (VFA) prediction model had a higher accuracy for MC3 (propionate, mol/100 mol) and MC4 (butyrate, mol/100 mol) than other volatile fatty acids. Among nutrients, NDFI (neutral detergent fiber intake) had a high effect on MC2 (acetate, mol/100 mol) and MC5 (valerate, mol/100 mol). In addition, in the case of MC3, MC4, and TVFA (total volatile fatty acid, mM), DMI (dry matter intake) was found to have the largest effect. Among the clusters of VFA for predicting methane, KM0 and KM2 changed according to MC2 and MC4, whereas KM1 was significantly affected by MC4 and TVFA. In the second experiment, predicted rumen volatile fatty acids and methane were used for the development of a carcass traits prediction model. The carcass traits prediction model showed that the prediction accuracy of the KNN model over the entire period was superior to those of other prediction models. Carcass characteristics in the growing period decreased in proportion to C2, (acetate, mM) and increased in proportion to C3 (propionate, mM) and C4 (butyrate, mM). The KNN''s R2 for the carcass trait prediction model in the growing period was as low as 0.21 to 0.29. All carcass traits in the early fattening period showed a positive relationship with C2 production. Also, among carcass traits, the error values were higher in the order of carcass weight (CWT), backfat thickness (BFT), eye muscle area (EMA), and marbling score (MSR). In carcass traits, C3 and C4 showed the greatest correlation with other metabolic variables during the late fattening period. During the entire fattening period, MSR showed a positive relationship with C4, C5, and methane. BFT showed the largest coefficient with C5, and EMA showed a large positive correlation with methane. CWT showed a positive regression coefficient for C3. On average, the R2 of KNN is 0.3, which means that up to 30% of nutrient intake could be predicted. The third experiment, a correlation between carcass characteristics, nutrient intake, VFA, and methane. As a result, it was confirmed that DMI, OMI (organic matter intake), CPI (crude protein intake), and TDNI (total digestible nutrient intake) had the greatest effect on carcass traits, and NDFI was also observed to affect carcass traits except for EMA. In the correlation with nutrient intake, only ADFI (acid detergent fiber) and TDNI showed a significant negative correlation (P<0.05). As for the carcass trait, CWT and BFT showed a high correlation, and EMA and MSR showed a strong correlation. The correlation between volatile fatty acids and carcass traits showed a positive correlation in TVFA, C2, and C4 for all carcass traits. Methane showed a significant negative correlation with carcass traits except for MSR and BFT, and the correlation with MSR was small. also, TVFA has a significant negative relationship with methane. Based on the median value of CWT, the high group and the low group were separated and compared. The contents of DMI and CPI in each stage of the specification were higher in the high group than in the low group in all periods. In the case of volatile fatty acids, C3 and C4 contents were higher in the high group than in the low group during the growing period, and C2 and C4 contents were higher in the early fattening period. The C2 content was high in the late fattening period. It was confirmed that the productivity of the conductor characteristics shown in this way was the best between CWT 480~500 kg in the high group. As for the proper nutrient intake, the average content of NDF and CP was 36.24% and 13.77% in the growing period, but 23.25% and 13.40% in the finisher. Also, 25.06% and 12.19% were observed in the late fattening period. Therefore, the TDN contents in the growing, early, and late fattening periods were 68%, 70%, and 74%. All figures were within the ranges indicated in the NRC specification standard. In conclusion, based on these results, it is necessary to secure data and upgrade the model for deriving the proper nutrient intake of Hanwoo steers. Sufficient nutrient supply of Hanwoo steers is important not only for the income of farmers but also for sustainable livestock production. Therefore, it is judged that the optimal nutrient intake of Hanwoo steers could be derived using the model.