This research focuses on modeling input data for simulation of construction operations. The first part of the research constitutes an evaluation and categorization of a number of real construction data sets. This categorization provides insight into the properties of construction data. In the second part of the research a number of methods are developed to enable accurate and efficient construction of input models when 1) observations are available, and 2) in a data deficient environment which is frequently encountered in construction planning. The third part of the research constitutes a sensitivity analysis of the effect of using different input models on the output of simulation. A number of realistic simulation models of construction processes were used as test beds for this part. A secondary objective of the research is the implementation of all methods and findings in portable software. The later part is documented in two main packages namely, BetaFit and BetaFitX.