Although imitation is more abundant and prevalent than innovation in firms' product and process development activities, it has been understudied in research on innovation and R&D management. For example, a valid and reliable objective firm-level measure of the intensity of imitation activity is lacking in the extant literature. This measure is necessary to understand the antecedents and consequences of firms' imitation activity, which has implications for R&D management. In this paper, we present novel methods that employ patent infringement litigations data to improve on the validity and reliability of measuring firms' imitation activity. We validate our proposed measure by presenting a first model and test of R&D as a multiple-output production function with R&D expenditure as the primary input, and innovation and imitation as joint outputs. This is in contrast to current R&D models as a single-output production function of either innovation or imitation. This study uses a sample of 227 public firms from the computer, semiconductor, and pharmaceutical industries in the United States during 1991-2010.