Current legislation in the United Kingdom stipulates that horses should not be slaughtered within sight of one another. However, abattoir personnel anecdotally report that, for semi-feral horses unused to restraint, co-slaughtering alongside a conspecific could reduce distress through social buffering and improve safety, but there is a lack of evidence to support this. CCTV footage from an English abattoir was assessed retrospectively with welfare indicators from when horses entered the kill pen until they were killed. Of 256 horses analysed, 12% (32/256) were co-slaughtered (alongside a conspecific) and 88% (224/256) individually. Co-slaughtered horses moved more in the pen, but individually slaughtered horses showed more agitated behaviour, required more encouragement to enter the kill pen, and experienced more slips or falls. Unrestrained horses (40%; 102/256) showed increased agitation, movement, and agonistic behaviour towards the operator and resisted entry to the kill pen compared to restrained horses (60%; 154/256). Positive interactions between conspecifics were seen in 94% (30/32) of co-slaughtered horses, and only 6% (1/16) showed a startled response to the first horse being shot, with a median time of 15 s between shots. This study highlights the impact that both conspecific and human interactions can have on equine welfare at slaughter. Semi-feral or unrestrained horses appear to experience increased distress compared to horses more familiar with human handling, and the presence of a conspecific at slaughter mitigated this.