In medical ophthalmology, refractive studies have become an integral part of a complete ophthalmic examination and second only to a slit-lamp biomicroscopic examination in determining visual function and ability. This study shows that the same ideology applies to refractive studies on dogs. The refraction technique has evolved in process, with refinement of the technology and methods used; with the development of handheld autorefractors utilised in paediatric refraction. Fifty guide dogs completed this study however five of the dogs were subsequently excluded and replaced because of poor compliance. Forty six of the dogs were Labrador Retrievers; one was a Golden Retriever, one a Great Dane cross, one a Labrador cross and one a cross breed. The median sample age was 17months with an interquartile range (IQR) of13 to 34months. We validated the agreement in refractive error measurement between the Welch Allyn ™ "SureSight" (WASS) autorefractor to the traditional Welch Allyn ™ handheld streak retinoscope (SR) by two experienced investigators, due to previous studies showing variability in measurement for both instruments. The refractive state for 60°/o of the guide dogs sampled at South African Guide Dogs Association (SAGDA) were emmetropic (-0.5 to +0.5D), 34% were hypermetropic (>+0.5D) with 6% myopic (<-0.5D). For agreement between the instruments, a wide range of differences using mean spherical equivalence (MSE) was observed between the 95% limits of agreement (-1.911 D to 1.698D). On average,measurements with the WASS were slightly lower compared to the SR (mean difference = 0.013). Both instruments showed a better average agreement in determining emmetropia with a tendency to underestimate refractive errors for greater negative and positive diopters. The WASS showed better agreement with the SR for refractive error measurement with higher reliability scale values (8 and above). Clinical astigmatism (>- 0.50 cylinder) was detected more readily on the WASS (37 dogs) than on the SR (9 dogs). Between investigators the MSE measurements differed significantly (P = 0.02), with the average agreement on the SR slightly better than the WASS. Overall, the handheld manual streak retinoscope remains the more practical, cost effective and efficient instrument for objective refraction in dogs. We can also conclude, that current subjective criteriae utilised in the selection of guide dogs at SAGA can be further improved by including objective methods of refraction. Copyright / Dissertation (MMedVet)--University of Pretoria, 2011. / Companion Animal Clinical Studies / unrestricted