This paper presents the evaluation of frequency-based approach predicted spring using acceleration signals that were collected from various road conditions. Random loadings in the forms of acceleration are nominal and more flexible for vehicle components fatigue assessment. In this analysis, the strain time history of the spring and acceleration signals of the suspension strut was measured from three different road conditions. The acceleration signals were then transformed into power spectra density (PSD). PSD cycle counter, like Lalanne, Dirlik, and narrow band approach, was applied to obtain equivalent load cycles. The stress response was obtained through having the equivalent load cycles with a spring modal frequency response function (FRF) and different stress criterion, like absolute maximum principal and critical plane approaches. Then, the stress response was used to predict the spring fatigue life using stress-life (S-N) approach. The results revealed that the harshest road condition was the rural road where the spring with fatigue life of 4.47 × 107 blocks to failure was obtained. The strain predicted fatigue life was used to validate the frequency-based predictions using a conservative approach. It was found that the Dirlik approach has shown the closest results to the strain life approach, which suggested that the Dirlik approach could be used for spring fatigue life prediction with the acceptable accuracy.