In many instances, a scientific examination of a painting can provide important information concerning attribution, provenance, and conservation. However, scientific examination is not a substitute for connoisseurship, and it is important for curators, art historians, and conservators to appreciate its potentials and limitations. This paper discusses some of the areas of application of science in the study of paintings, the procedure to follow in requesting an analysis by Analytical Research Services (CCI), and the necessity of having reliable data banks of information on artists' materials. Several examples are provided of the scientific examination of paintings: The Virgin and Christ Child by Bernard van Orley; Two Water-Mills by Meindert Hebbema; a Self Portrait by Henry H. Cross; and a double painting, Yan Mortuary Poles, by Emily Carr. The study of an early map of the Eastern Canadian Arctic is also described. The paper concludes with a discussion of the limitations of scientific examinations.