In an experiment conducted in NE Spain, cereals and legumes were rotated for seven cycles using two different tillage techniques – conventional tillage and direct drilling (no-tillage: NT). Straw was removed after harvesting in both tillage systems. With NT, more than 30% of the soil surface was still covered by residues (stubble) at sowing. The soil was Calcic Cambisol and the climate aridity index was 0.76. The climatic water balance was negative, at -187 mm y-1 from 1950 to 1980. The aims of the study were to determine the effect of tillage system on 2-D macroporosity by micromorphology, and bulk density (excavation method) of the upper layer of the ploughed horizon. To identify differences, either parametric or nonparametric statistical tests were performed depending on sample size. Direct drilling with residue removal affected bulk density, macroporosity and mean macropore area in the top 10 cm of the profile. The same effects were observed under conventional tillage. The upper layer of the studied horizon had higher total porosity than the layer beneath with both treatments. The upper layer of the horizon showed unfavourable physical conditions with direct drilling.