In groups of gestating sows, accessing feed supply may be a source of tension and induce aggressiveness and fights. Using electronic sow feeders (ESF) could help to ensure that each sow eats its entire ration in a grouphoused pen. The objective of this study was to evaluate the impact of a competitive feeding situation related to the management of the ESF, on the overall behaviour, physical activity, and energy requirements of gestating sows. A total of 32 sows, divided into two groups, were involved in the study. The feed competition was created by closing the access of one of the two feeders available in each gestation room for 5 consecutive days. This situation was repeated twice, each time preceded by one control week during which both feeders were available. Automatons, accelerometers and manual video analysis recorded each sow's feeding, drinking and social behaviour, their physical activity, occupation and location in the pen. Linear mixed effect models were used, including the fixed effects of treatment (feed competition or control), repetition (first or second period), group (2 modalities), social ranking (3 modalities), their interactions, and the random effect of sow. During competitive feeding situations, the number of negative interactions doubled (P < 0.001), the sows spending more time standing in the morning (P < 0.001) and in the areas located further away from the feeder (P = 0.005). The number of and time spent for nutritive visits were not affected by treatment, whereas the number and duration of non-nutritive visits decreased (P < 0.001) during feed competition. This later effect was more marked in low-and high-ranking sows (interaction, P < 0.01). During feed competition sow spent less time sleeping (P < 0.01) and more time exploring the feeder (P < 0.001), these effects being less marked in period 2. The calculated metabolizable energy (ME) requirement, and therefore the quantity of feed required, increased during feed competition (P < 0.001), the effect being greater during the second period than the first (interaction P = 0.04). Moreover, ME requirement was greater for higher compared to lower-ranking sows (P < 0.001). The treatment had no significant effect on body weight and back-fat-thickness, neither on health criteria. To conclude, these results indicate that short-term dysfunction of ESF affect the overall behaviour of sows and increase negative interactions and energy requirements, without affecting their performances or health.