In tunnelling and in other underground constructions, a detailed knowl
edge of the rock mass to be excavated is essential for reasons of safe
ty, reinforcement need and cost and time schedule for the excavation.
This paper presents a new approach to how drill performance parameters
can be used for Rock Quality Designation (RQD) prediction. Prediction
s are based not only on traditional parameters as penetration rate and
torque, but also on the variability of each parameter which shows a c
lose correlation, to fracturing. Since predictions are based on drill
monitoring data with high resolution, the prediction can provide detai
led information of the structural geometry of the rock mass ahead of t
he operation. Examples are given from a railway tunnel in northern Swe
den, built in granitic bed-rock.