Experimental measurements of the loading and wake of horizontal axis tidal stream turbines are required to inform the development of numerical models of turbine interactions. Several constraints and scaling considerations influence thedesign of a scale-representation of a tidal stream rotor. These include the conflicting requirements of Froude and Reynolds number as well as blockage scaling. Due to the dimensions of existing tank testing facilities, the Reynolds number based on the blade chord is usually significantly reduced compared with fullscale tidal turbines. At such reduced Reynolds numbers, the combined effect of decreased lift and increased drag leads to much lower torque and hence the variation of both power coefficient and thrust coefficient are different to those of a full-scale rotor. If the blade geometry is modified, rather than geometrically scaled, blade sections can be employed that demonstrate a significantly better performance at low Reynolds number. This paper presents several arguments for modifying the geometry of model scale rotors for the purposes of representative wake modelling. Performance predictions for 3-bladed rotors are obtained using Tidal Bladed, a rotor design tool based on a blade element momentum (BEM) method. Experimental testing is a key component of the Energy Technologies Institute commissioned PerAWaT project. Case studies are presented for rotors of approximately 1/30th and 1/70th of a representative full-scale rotor. It is shown that this rotor design process can be used to produce a variation of thrust coefficient with tip-speed ratio that is similar to a full-scale device.