The volume reduction behaviour of powders has been quantified by means of the 'in-die' yield pressure (YP) using Heckel analysis. However, because different YPs are reported for the same material, the experimental conditions influencing this material-constant were investigated. Silicified microcrystalline cellulose was compressed into flat-faced and convex tablets using a compaction simulator instrumented with load and displacement transducers. During compression, upper and lower punch force and displacement data were recorded and corrected for punch deformation. A symmetrical triangle wave compression profile was used and the instantaneous punch velocity was kept constant (5mm/s). Individual tablet height and weight were used for Heckel analysis. The influence of the 'effective compression pressure' (P(EFF)) (ranging from 10 to 350 MPa), punch diameter (PD) (4, 9.5 and 12 mm) and filling depth (FD) (4.5, 7.5 and 10.5mm) on YP was statistically evaluated using Response Surface Modelling software. A quadratic surface response equation, describing the relationship between P(EFF), PD, FD and YP, was proposed for concave (Adj R(2): 0.8424; S.D.: 14.60 MPa) and flat-faced (Adj R(2): 0.8409; S.D.: 4.49 MPa) punches. YP and tensile strength were mainly determined by P(EFF), irrespective of punch curvature. FD and PD had only a minor influence on the YP, although more pronounced for the concave punches. The method used resulted in reproducible P(EFF) and tensile strength values and the flat-faced tablets showed less weight variation. Flat-faced punches are preferred over punches with a concave surface when investigating the volume reduction behaviour of a powder by means of Heckel analysis and the experimental parameters should be reported.