The present paper aims to provide a benchmark set of sail pressure distributions for the validation of numerical codes. Modern upwind sails were built at 1/15th-scale and tested in a wind tunnel. The sails were built as a fibreglass sandwich and the pressures were transmitted through channels inside the core. The sails were supported with wind-transparent fishing lines, and a flat plate was used to model the boat deck. Four genoa and four mainsail trims were tested, resulting in 16 test conditions. The geometry of the sails and the pressure coefficients along four horizontal sections of each sail are provided herein. The predicted pressures from a vortex lattice method numerical code for the various sail test configurations were compared against the experimental data to verify the capability of the present database of providing a valuable benchmark. There was generally good agreement between the computed and measured pressures. The differences can mainly be explained by the action of viscosity. For example the numerical code could not model a leading edge separation bubble that occurred for some of the “tight” trims, and so the computations gave high leading edge suctions, and moved the second suction peak forward compared to the measured pressures.