The only serious limitations of field growth studies are due to the erratic nature of the curves caused by short-term weather variations and the less apparent deviations caused by secular changes. The method of constant-fitting is applied, via a suitable mathematical model and experimental design, to separate growth and weather effects. A detailed account of the analysis is given, using a trial conducted as a test of the technique as an example, and the future possibilities of the technique in field growth studies are discussed.