Partial-face rock excavation machines, which are equipped with conical picks form a major share in the mineral production and in the roadway excavation for underground mines. Consequently, the performance of these machines plays a key role for the success of the project where the prediction of its performance generally relies on the instrumented cutting tests requiring a large rock sample. These rock samples, however, are in most cases impossible to be collected from excavation sites, especially from underground mines due to safety issues and harsh working environment. Therefore, alternative reliable methods requiring a smaller amount of rock sample are preferred. This paper was tailored to address a new plot to convert the index cutting data from the core cutting test into the optimum rock cutting data where these data are used for performance prediction and machine design. A special emphasis was given to medium strength rocks in which partial-face rock excavation machines frequently operate. Nine rock samples were collected from quarries and roadheader excavation sites, which have a strength spectrum from 1.89 to 99.92 MPa and these samples were subjected to rock cutting tests both using the procedure for the core-cutting test in single pick cutting mode and for the conical pick cutting test in relieved mode. In virtue of cutting tests, predictive plots were developed to determine the optimum conical pick cutting data through conducting a single pick cutting test using a small block sample. The major conclusion drawn from the experimental results is in the absence of large rock sample for an instrumented cutting test, small block or core samples can be utilised for estimation of optimum rock cutting data, especially for very low and medium strength rocks.