In this study we used a new call feeding station, which enabled sows to learn that they have access to feed only after an individual acoustic signal was given. We tested whether this call feeding station is able to reduce agonistic interactions between sows and whether effects of call feeding can further be improved by enrichment. A total of 85 gestating sows were kept in a dynamic group in a large waiting area (207 m²) equipped with littered laying areas and an outside area. During a control treatment sows were fed in a normal electronic feeding station once a day (NF1-). Before testing the call feeding station sows had been conditioned for an acoustic signal (a trisyllabic "name") and learned that they were allowed to enter the feeding station only after their name was called. In the call feeding station sows were fed either once (CF1-) or twice a day (CF2-). In addition, we tested for effects of further enrichment such as straw in the activity area (CF2+). Agonistic behaviors and number of sows were observed by video in continuous recording from 0600 to 1800 h in an area (4 by 3.25 m) in front of the feeding station in periods of 4 d (NF1- = 7 periods, CF1- = 5 periods, CF2- = 3 periods, and CF2+ = 4 periods) and analyzed using mixed models. During each observation period sows were scored for wounds and body lesions at different body parts. From 0600 to 1100 h the proportion of agonistic interactions was much greater in the feeding mode NF1- compared with CF1- (feeding mode × time of day: P < 0.001) and in CF1- agonistic interactions were on a low level throughout the whole feeding cycle. The feeding frequency and the additional presence of straw in the activity area did not affect the proportion of agonistic interactions (all P > 0.05). The results on the number of sows in front of the feeding station mirrored the findings for agonistic interactions. In NF1- more sows were involved in agonistic interactions compared with CF1- (83.1 ± 12.9% vs. 61.5 ± 19.6%; P = 0.005) but there was no difference between the call feeding station (CF) treatments. The number of severe lesions was greater at the head (P = 0.014) and the flank (P = 0.006) but not at the shoulders (P = 0.057) and the hindquarter (P = 0.426) in NF1- compared with CF1-. The results suggest that signaling the feeding time individually increases the predictability for access to the feeding station and consequently reduces competition between sows.