The recent development of various aerospace applications utilizing Ni-rich NiTi Shape memory Alloys (SMAs) as actuators motivated the need to characterize the cyclic response and the transformation fatigue behavior of such alloys. The fatigue life validation and certification of new designs is required in order to be implemented and used in future applications. For that purpose, a custom built fatigue test frame was designed to perform isobaric thermally induced transformation cycles on small dogbones SMA actuators (test gauge cross-section up to: 1.270 x 0.508 mm2). A parametric study on the cyclic response and transformation fatigue behavior of Ni-rich NiTi SMAs led to the optimization of several material/process and test parameters, namely: the applied stress range, the heat treatment, the heat treatment environment and the specimen thickness. However, fatigue testing was performed in a chilled waterless glycol environment maintained at a temperature of 5°C that showed evidence of corrosion-assisted transformationfatigue failure. Therefore, it was necessary to build a fatigue test frame that would employ a dry and inert cooling methodology to get away from any detrimental interactions between the specimens and the cooling medium (corrosion). The selected cooling method was gaseous nitrogen, sprayed into a thermally insulated chamber, maintaining a temperature of -20°C. The design of the gaseous nitrogen cooling was done in such a way that the actuation frequency is similar to the one obtained using the original design (~ 0.1 Hz). For both cooling methods, Joule resistive heating was used to heat the specimens. In addition and motivated by the difference in surface quality resulting from different terial processing such as EDM wire cutting and heat treatments, EDM recast layer and oxide layer were removed. The removal was followed by an ultra-fine polish (0.05 μm) that was performed on a subset of the fatigue specimens. Experimental results are presented for full actuation of the SMA actuators and are given in terms of applied stress, accumulated plastic strain and number of cycles to failure. In addition, the assessment of the influence of the surface quality is supported by fatigue tests results and post-failure microstructure analysis. © 2009 SPIE.