Residual stresses appear in biological tissues as a consequence of growth and remodeling processes under physiological or pathological conditions. They are defined as the stresses present in a material in the absence of external loads or body forces [1], and can be directly computed based on existing growth laws within the framework of continuum nonlinear mechanics. However, the phenomenological nature of such models, usually involving a non-negligible amount of parameters (being some of them unmeasurable), limits its applicability in tissue biomechanics. To address this, we propose a family of inverse physics-guided algorithms to estimate nonlinear residual stresses, that use input information extracted from conventional biomechanical tests. Notably, experimental data needed from the referred experiments are domain displacement fields, which can be obtained in the course of the assays by means of digital image correlation (DIC) techniques. The framework is evaluated through different examples related to the recovery of residual stress fields induced by tissue growth, across varying levels of noise in the input data-set. Results show that accurate reconstructions of complex residual stress patterns are possible, even for high errors in the input displacement field. The approach is demonstrated to be independent on the intrinsic mechanisms and underlying biophysics of growth, and can be used to get residual stress fields alternatively to growth modeling. Also, the proposed scheme can be used as a tool to validate growth models based on experimental data.