The feed barrier has been shown to have major effects on feeding and social behaviour of housed dairy cows. Computerised systems allow continuous monitoring of feeding and drinking behaviour and are frequently used in research. However, the design of these feeders can obscure the cow's visual field. The objective of the study was to determine if obscuring the cows' visual field whilst feeding would increase vigilance behaviour and alter normal feeding behaviour. Twenty-four Holstein cows were divided into 4 groups. Two pens had headlock feed barriers (a conventional design where the cow puts her head through the bars of the barrier to reach food) and 2 pens had an electronic system (INSENTEC, Holland) where each cow is fitted with a passive transponder attached to the ear tag which permits access to feed when the transponder is detected. Groups were exposed to four treatments (6, 5, 4, or 3 feeding spaces) following a Latin Square design and repeated on each feed barrier. Restricted Maximum Likelihood (REML) mixed models were fitted to data. Average daily feeding time was higher (P<0.05) when cows were fed using the headlock system (265.8 ± 9.6; mean ± SE) compared to the electronic feeding stations (224.2 ± 7.01; mean ± SE). All groups of cows displayed vigilance scans, however, type of barrier did not have a significant difference on frequency of scans (headlock system, 91.07 ± 6.34; electronic system, 104.64 ± 7.10), and neither did number of feeding stations or dominance rank. The low status cows did, on average, scan slightly more times than the high status cows (however, this was not significant). These results indicate that neither feeder design nor stocking rate affected vigilance in dairy cows, at least over the treatment conditions assessed in the current study.