The design of mining excavations in rock requires access to a representative geotechnical model that includes the mechanical properties of the rock mass. The available geotechnical data provide the necessary input to analytical, numerical, and empirical design tools. Consequently, any geotechnical analysis is influenced by the quality of the input data. Therefore, to understand and mitigate the design risk caused by data uncertainty, it is critical to evaluate the level of confidence in the collected geotechnical data. A practical limitation of current mine design practice is the absence of quantitative guidelines to select the number of laboratory tests required. This investigation employs small-sampling theory to determine the minimum number of tests necessary to obtain predefined confidence intervals in intact rock estimates at South African mines. A key element of this work is the introduction of geotechnical domain complexity as a significant factor in establishing quantitative recommendations for the required minimum number of laboratory tests. A tangible contribution of this work is the development of an original methodology for planning laboratory testing campaigns for a new mining project or for updating the geotechnical database of operating mines. The proposed quantitative methods can eventually replace subjective assessments in addressing data collection requirements.