The problem of sensation of draught in ventilated spaces is connected to inappropriate velocities in the occupied zone. In Scandinavia, velocities higher than 0.15 m/s are said to be an indicator of that occupants are likely to feel discomfort. Therefore knowledge of the flow field (both mean velocities and fluctuations) is necessary. Both experimental and numerical analysis of the flow field in a full scale room ventilated by a slot inlet, with two inlet Reynolds numbers 2440 and 7110, have been carried out . Results from both approaches show that the location of the maximum velocity near the floor is nearly independent of the Reynolds number. For a two-dimensional room, the maximum velocity at the floor level occurred at about 213 room length from the supply. The distance from the floor level is dependent on the inlet Reynolds number. The velocity profiles far away from the wall opposite to the inlet device have the same character as a wall jet profile. However, close to the corners they are transformed. The relative turbulence intensities measured in the return flow region are questionable, because of a hot wire's inability to record large fluctuations at low mean velocities. These turbulence intensities close to floor level vary from 15 to 80 % and as the authors have pointed out previously hot wires do not indicate the real value of the turbulence intensities beyond 20%. Difficulties appear in numerical predictions of return flow properties. Comparison between predicted values and experimentally obtained values show a reasonable agreement. This is promising for future CFD-predictions. However, there is a need for an appropriate measurement technique that can cope with reversing flow.