This book is intended to be a convenient source of ideas and methods (a companion) for practicing cost analysts. It deals with five areas: a global perspective of cost analysis and estimating as applications of system identification; the nature and use of cost-progress curves; the use and development of cost-estimating relations; the application of Rayleigh methods for analyzing development program costs; and, the implementation of engineering and statistical principles when estimating operating and support costs. Some of the material, like the introduction to cost-estimating relations, is elementary. These parts probably will be useful mostly to people who have limited experience in cost analysis. Other parts, like the discussion of system identification as a paradigm for cost analysis and the section on using physical principles to make cost-estimating relationships, are intended to be interesting and useful to the experienced practitioner. The approach in each of the five areas of consideration is to base cost analyses on principles of economics, physics, and technology, and to develop them with careful applications of mathematics and statistics. Working this way may seem unnecessarily fussy. The complex situations cost analysts must treat usually far exceed what can be modeled completely from first principles. Almost always, cost estimates and analyses need inputs from experience, known as "analysts' judgments." Since judgment will always play an essential role, why bo