In this investigation, diamond-shaped samples laser-cut from nitinol tubing were used to investigate the effect of variable amplitude fatigue loading. Alternating strain amplitudes of 0.30 %, 0.35 %, and 0.40 % were applied with periodic high strain amplitude of 0.70 %. The high strain was applied one time for every 40 low strain cycles. Using the obtained data, the appropriateness of the Palmgren-Miner’s linear and the Marco-Starkey nonlinear cumulative damage models was investigated. The variable amplitude testing results indicated a trend of decreasing fatigue life with the addition of periodic high strain. Data obtained were not sufficient to demonstrate that the Palmgren-Miners model was an accurate predictor. However, the Markco-Starkey model was found to provide additional flexibility in obtaining more precise predictions. With fatigue-to-fracture data and an appropriate variable amplitude model, the life associated with the variable amplitude loading can be estimated, and burdensome and lengthy physical variable amplitude testing can be avoided.