In this paper, the theoretical basis for construction labor productivity me
asurement is presented. In particular, the theoretical basis for baseline p
roductivity measurements is developed by examining a productivity database
consisting of 23 projects involving masonry construction. An important hypo
thesis is presented showing that as the design becomes more complex, the ba
seline productivity worsens. It is also hypothesized that higher values of
the coefficient of variation indicates a higher variability in management a
nd craft skills and in the use of technology. Two measures are proposed to
measure the performance of individual projects: The disruption index and th
e project management index. These two measures identify the best and worst
performing projects. Cumulative probability distributions of the disruption
index and the project management index were also developed to evaluate the
23-project database and compare it with other databases. The hypotheses de
veloped from the 23-masonry project database were tested against an 8-proje
ct database of concrete formwork and a 12-project database of structural st
eel erection. Strong support for each hypothesis was found using the two ad
ditional databases.