Background: A disturbing number of ridden horses are lame, although assumed by riders to be pain free. A Ridden Horse Pain Ethogram (RHpE) was developed; an RHpE score of ≥8/24 is likely to indicate the presence of musculoskeletal pain. Pain assessment in Icelandic horses is challenging; the RHpE may facilitate lameness recognition. Objectives: To determine the repeatability of application of an adapted RHpE to Icelandic horses by a single observer. To evaluate the relationship between RHpE scores and lameness or abnormalities of canter. Study design: Prospective, cross-sectional study. Methods: Video-recordings of 30 Icelandic horses performing a standardised test (walk, trot, tölt, canter) were acquired. The adapted RHpE, to include tölt, was applied to all recordings twice in random order by a single observer. Agreement was assessed using Cohen's kappa. The presence of lameness, abnormalities of canter and lameness grade was determined by an independent expert. The association between maximum lameness grade and RHpE score was tested using Spearman's rank correlation. Results: All horses exhibited both lameness, the majority in >1 limb, and abnormalities of canter. The RHpE score was ≥8/24 in 96% (median 10/24 and 10.5/24, respectively, for the two assessments). There was substantial to near-perfect agreement for the majority of behaviours (63.6%) and moderate or fair agreement for 32%. There was no correlation between maximum lameness grades and RHpE scores. Main limitations: The absence of non-lame horses with correct saddle fit for horse and rider. Conclusions: There was good repeatability of application of the RHpE. Musculoskeletal pain was associated with RHpE scores of ≥8/24 in the majority of horses. Data acquired from non-lame horses, ridden in well-fitting tack by riders of appropriate size, are needed to determine if the threshold of step frequency in trot needs to be adjusted for more accurate application of the RHpE in Icelandic horses.