This research consisted in comparing the fatigue crack growth (FCG) performance of four HSLA/ AHSS steels used in automotive applications and with different microstructures, and the application of some prediction models for the da/dN versus ΔK traditional sigmoidal curve as a function of the R load ratio. FCG tests were carried out on C(T) test specimens with R-ratios varying between 0.03 and 0.7. Using the original and empirical methodology proposed by Paris and Erdogan to describe the da/dN-ΔK relationship, the results showed significant differences in function of microstructure, and a deleterious effect of R-ratio increase on the crack growth rate. In order to check existing methodologies based on physical considerations for predicting the fatigue behavior of materials and the effect of the R-ratio mainly in the fatigue threshold ΔK th region, the well-known crack closure model proposed by Elber, an approach using two parameters as a driving force for the crack growth proposed by Vasudevan and co-authors and a combination of these two models recently proposed by Zhu and co-authors were compared. The manifestation of crack closure and its qualitatively expected dependence on the R-ratio were verified for the studied steels, but the Elber model was not able to provide a master curve that accurately summarized the effect of the R-ratio on the sigmoidal fatigue curve of steels. The combined use of two critical thresholds, ΔK th * and K max *, for predicting fatigue crack growth according to the Vasudevan model also did not provide accurate results in evaluating the effect of the R-ratio. Regardless of the verified dispersions, there is a connection between the two-parameter methodology and crack closure, hence the model by Zhu and co-authors could be a promising alternative. However, this model also showed significant dispersions and was unable to create a master curve to adequately predict the effect of R-ratio on crack growth. Thus, it can be concluded that this research topic is still open, requiring a more in-depth phenomenological knowledge to predict the effect of the R-ratio on FCG.