The effects of terminal sire line, halothane genotype and age on feeding patterns and growth performance were studied in 208 castrated growing-finishing pigs distributed in two batches. In batch 1 (no. = 130), pigs came from crosses of NN Landrace × Large White sows with a Pietrain Nn (Pi-Nn) or a Large White × Pietrain Nn terminal sire line (LwPi-Nn). In batch 2 (no. = 78), the same sows were used but crossed with three different terminal sire lines : a Pietrain nn (Pi-nn), and two Pietrain NN (Pi-NNa and Pi-NNb). Growth performance and feeding patterns of the progeny (Nn or NN for the halothane gene) were measured from 67 to 166 days of age. Pigs were housed in groups of 10 and 13 individuals in Batch 1 and 2, respectively, with a random mixed-breed and halothane genotype sample in each group (space allowance 1·3 m2 per pig and 1·1 m2 per pig, respectively). Feeding patterns were monitored with a computerized food intake recording system (IVOG®-station) and every 3 weeks pigs were weighed and backfat and loin-muscle depth were ultrasonically recorded (PIGLOG®). Carcass quality was assessed with the Fat-o-Meater grading probe. In batch 1, halothane genotype did not have a significant effect on any of the feeding patterns recorded, but Nn individuals had a significantly higher body weight (P <0·05), loin-muscle depth (P <0·05) and lower backfat thickness (P <0·01) in the last measurement taken, 3/4 carcass loin depth (P <0·05) and lower carcass last rib backfat (P <0·05) than NN pigs. Terminal sire line had a significant effect on all feeding patterns recorded except for feeding rate, Pi-Nn sired pigs showing a significantly higher food intake per visit (P <0·05) and feeder occupation time per visit (P <0·05) and lower number of visits (P <0·001) compared with LwPi-Nn sired pigs which, in turn, showed significantly higher food intake per day (P <0·001) and feeder occupation time per day (P <0·01). Terminal sire also affected growth performance and body composition, Pi-Nn sired pigs having a significantly lower body weight (P <0·001) and backfat thickness (P <0·001) but higher killing-out proportion and 3/4 loin depth (P <0·01) than LwPi-Nn sired pigs. In batch 2, Pi-NNb sired pigs showed a significantly lower food intake per day compared with the progeny of the other two terminal sires lines at some of the age measurements taken (P <0·05). The rest of the feeding patterns was not affected by terminal sire line. Body and carcass weights were also significantly lower (P <0·01) for Pi-NNb sired pigs, but their killing-out proportion was higher than Pi-NNa sired pigs. With regard to these variables, Pi-nn sired pigs held an intermediate position between the two NN terminal sire lines. In both batches, age was associated with a significant increase in food intake per visit and per day and feeding rate (P <0·001) and a decrease in feeder occupation time per visit and per day and frequency of visits to the feeder (P <0·001). Overall, the present results suggest differences between terminal sire lines for feeding patterns and confirm their evolution with age from short and frequent meals to long and larger ones in growing-finishing pigs. Under our conditions, the effects of terminal sire line on feeding patterns and growth performance surpassed those of the halothane genotype.