Today the prevailing view in child psychiatry is the biopsychosocial model of child development, but this was not always the case. Prior to World War II, and even in the post-war period, the environment was considered the major determinant of the child's development; the role of the child's constitutional characteristics was overlooked. The theory of temperament, formulated by the author and her colleagues, holds that these constitutional characteristics are important in the child's development and interact with the environment. We tested the theory of temperament through a major longitudinal study, which required innovative approaches to data collection and analysis. Data obtained from the 133 children and accumulated over the past 35 years have been described in earlier publications. This paper traces the evolution of our methodology from the original concept to qualitative data collection and quantitative measurement. The methodological and theoretical dilemmas encountered in our research and some of the implications of our findings are also discussed.