The paper provides a practical procedure to estimate the chemical composition of pigs, their compositional growth and the expected feed intake from measurements of body weight (BW) and backfat thickness (P2) serially performed in vivo. A farm data set provided information on 920 individuals including BW, measured at 71 ± 4 (t1), 126 ± 5 (t2) and 184 ± 5 (t3) days of age, of P2 at t2 and t3, and of voluntary daily feed intake (FI), recorded over the period from t2 to t3 by automated IVOG feeders. Body lipid mass was estimated as L= (9.17 + 0.70*P2) *BW/100 and the other chemical constituents were predicted from fat free empty body mass using Gompertz growth functions and allometry. Using individual changes of body composition from age t2 to t3, energy requirements for maintenance and growth and the corresponding predicted feed intakes (PFI) were estimated. Measured FI were analysed for the effects of month, batch (within month), BWt2, P2t2, average metabolic weight, average daily gain and variation of P2 from t2 to t3. The same model was run again replacing the direct simple body measurements (BW and P2) with the estimated values of PFI as source of variation. Results. The Gompertz estimates of mature protein mass (Pm), relative growth rate parameter (B) and lipid to protein ratio at maturity were 43.5 ± 5.8 kg, 0.0116 ± 0.0011 d-1 and 1.81 ± 0.30, respectively. The current protein mass averaged 18.5 + 1.6 kg and the daily retentions of protein and lipid were 177 ± 21 and 239 ± 62 g/d, respectively. FI and PFI averaged 2.824 ± 0.448 and 2.814 ± 0.393 kg/d, respectively. In the ANOVA of the FI data, the replacement of direct body measurements by PFI did not change the proportion of variance explained (83%) and the RSD (0.199 g/d). The two sets of residual feed intake values obtained from the two ANOVA were highly correlated (RSD = 0.043 kg/d; R2= 0.961). Agreement between predicted and determined feed intakes provided a reasonable guarantee to the estimated (based on BW and P2) changes of body composition. Thus, a scheduled protocol of measurement of BW and P2 over the course of growth, coupled with the use of allometry, can be proposed to estimate in vivo the change of the chemical status of pigs kept under non limiting conditions.