The objective of this paper was to explore the potential of hybrid-polarized (RH and RV) RISAT-1 SAR data to retrieve the height of wheat crop—an important winter crop in South Asian countries including India. The images acquired over north-west India in 2015 covered critical growth stages of wheat. The field campaigns were carried out in synchronous with the SAR passes. Considering the dominant role of underlying soil cover in the total backscatter $({\bm{\sigma }_{\bm{total}}^0})$ response from a target, we propose that refining the $\bm{\sigma }_{\bm{total}}^0$ by reducing the effect of underlying soil can significantly improve the retrieval accuracy of crop height (CH). To achieve this, we modified the existing water cloud model (WCM) to estimate soil-corrected vegetation backscatter $({\bm{\sigma }_{\bm{veg}}^0})$ . Leaf area index and interaction factor showed great potential as the vegetation descriptors in modeling $\bm{\sigma }_{\bm{total}}^0$ using WCM. A comparative analysis between the CH retrieved from $\bm{\sigma }_{\bm{total}}^0$ and $\bm{\sigma }_{\bm{veg}}^0$ using multilayer perceptron neural networks revealed the response of C-band backscatter to CH. CH was moderately correlated to $\bm{\sigma }_{\bm{total}}^0$ , but the results improved considerably with the substitution of $\bm{\sigma }_{\bm{total}}^0$ with $\bm{\sigma }_{\bm{veg}}^0$ . This holds true particularly in the early growth stages of crop growth when the vegetation cover is scarce and there is a substantial effect of soil background on the remote sensing signal. Thus, the results suggest suitability of C-band hybrid-polarized data for the assessment of CH.