The effect of measurement scale on hydraulic properties close to saturation was investigated in a clay soil. Results from measurements on undisturbed “standard” small soil core samples were compared with results from three large intact soil monolith samples which were assumed to reflect natural soil hydraulic behaviour. An intermediate sample size, which was used to characterize soil water retention relations (θ(ψ)) in each layer in the intact monoliths, was obtained by cutting the intact monoliths in three layers. The small cores were subsequently sampled from the cut monoliths so that all measurements were made on the same soil material. Measurements of θ(ψ) and saturated hydraulic conductivity ( K s ) were compared, as well as the exponential relationships between K s and macroporosity ( ϕ ma ), the latter derived from the θ(ψ) data for small cores and from specific yields for intact monoliths. Between soil water pressure heads of −60 cm and −15 cm the small core data and the cut monolith data showed similar θ(ψ) relations, whereas they diverged closer to saturation. When comparing K s and ϕ ma , small core data were correlated with the intact monolith data although the small cores had smaller values. Thus, the “standard” small cores can satisfactorily predict the investigated soil hydraulic properties for a natural soil at pressure heads equal or less than −15 cm, but they may be in error, especially in soils with macropores, in the pressure head range −15 cm to saturation, i.e. in the largest macropores.