This article examines some of the difficulties associated with applying experimental results to field conditions, especially in the area of concrete repair. Existing research and testing methods used for evaluating the performance of a concrete repair are unsatisfactory, with the laboratory test results often shown to be inconsistent. These results are unreliable because the tests are related to concrete produced and cured in the laboratory. This does not allow for a complete understanding of the material's in-place behavior and its effects on repaired structures. Instead, researchers should study repair-related issues of concrete under field conditions. The environment, repair location, geometry, restraint and nonuniformity should be considered before a prognosis of performance and longevity is made. It is recommended that the industry test cementitious repair materials in reproducible ways so that practitioners are confident when specifying and using them. If the repaired structure is to be durable, appropriate measures need to be taken to control volume changes and the resulting induced cracking of the cementitious composites used for repair. Testing of related fundamental properties that control the durability of concrete have to be perfected to allow for reliable prediction of performance in the actual service environment.