Aims: To determine improvements in third metacarpal (Mc3) condylar microanatomy attributable to preconditioning exercise. To investigate developmental causes of Mc3 condylar fracture. Methods: Twelve Thoroughbred horses were raised at pasture; six received preconditioning exercise from 10 days. Calcein labels were administered 19 and 11 days prior to euthanasia at 18 months. Six horses also received 2 seasons of race-training and were euthanised at 3 years. Slices were taken from the distal Mc3 condyle in the frontal and dorsal- and palmar-oblique frontal planes, scanned with DXA and macerated (frontal slices) or embedded in PMMA (oblique slices). Articular calcified cartilage (ACC) and subchondral bone (SCB) in oblique slices were imaged using confocal scanning light microscopy and quantitative backscattered electron scanning electron microscopy. ACC and SCB in the palmar slice lateral parasagittal grooves were imaged using µCT and nanoindentation tested. Results: Characteristic spatial variations in ACC and SCB histomorphometric parameters were present, none of which was significantly related to preconditioning exercise. Thickened, aberrantly mineralised ACC was found in 13/24 parasagittal grooves in the palmar slices and on the sagittal ridge of 4/12 dorsal slices of 18-month-old horses. Deep to thickened ACC, SCB had an open marrow structure, having not adopted the buttress morphology of the normal SCB plate. SCB in 3-year-old horses had incorporated early ACC defects as notches in parasagittal grooves and a hyaline cartilage island in a sagittal ridge. ACC was less stiff and SCB more stiff in affected than unaffected parasagittal grooves. Chondroclastic resorption in the parasagittal groove may be retarded as early as 3-6 months, possibly due to localised inhibition of ACC mineralisation. Linear defects in the Mc3 parasagittal groove may develop prior to entry to race training and are not significantly affected by preconditioning exercise. Early identification of affected individuals should aid in reducing condylar fracture risk.