Cartilage canals have been shown to contain discontinuous blood vessels that enable circulating bacteria to bind to cartilage matrix, leading to vascular occlusion and associated pathological changes in pigs and chickens. It is also inconsistently reported that cartilage canals are surrounded by a cellular or acellular wall that may influence whether bacterial binding can occur. It is not known whether equine cartilage canals contain discontinuous endothelium or are surrounded by a wall. This study aimed to examine whether there were discontinuities in the endothelium of cartilage canal vessels, and whether canals had a cellular or acellular wall, in the epiphyseal growth cartilage of foals. Epiphyseal growth cartilage from the proximal third of the medial trochlear ridge of the distal femur from six healthy foals that were 1, 24, 35, 47, 118 and 122 days old and of different breeds and sexes was examined by light microscopy (LM), transmission electron microscopy (TEM) and immunohistochemistry. The majority of patent cartilage canals contained blood vessels that were lined by a thin layer of continuous endothelium. Fenestrations were found in two locations in one venule in a patent cartilage canal located deep in the growth cartilage and close to the ossification front in the 118‐day‐old foal. Chondrifying cartilage canals in all TEM‐examined foals contained degenerated endothelial cells that were detached from the basement membrane, resulting in gap formation. Thirty‐three percent of all canals were surrounded by a hypercellular rim that was interpreted as contribution of chondrocytes to growth cartilage. On LM, 69% of all cartilage canals were surrounded by a ring of matrix that stained intensely eosinophilic and consisted of collagen fibres on TEM that were confirmed to be collagen type I by immunohistochemistry. In summary, two types of discontinuity were observed in the endothelium of equine epiphyseal cartilage canal vessels: fenestrations were observed in a patent cartilage canal in the 118‐day‐old foal; and gaps were observed in chondrifying cartilage canals in all TEM‐examined foals. Canals were not surrounded by any cellular wall, but a large proportion was surrounded by an acellular wall consisting of collagen type I. Bacterial binding can therefore probably occur in horses by mechanisms that are similar to those previously demonstrated in pigs and chickens.