Rock-physics models are used increasingly to link fluid and mechanical deformation parameters for dynamic elastic modeling. We explore the input parameters of an analytical stress-dependent rock-physics model. To do this, we invert for the stress-dependent microcrack parameters of more than 150 sedimentary rock velocity-stress core measurements taken from a literature survey. The inversion scheme is based on a microstructural effective-medium formulation defined by a second-rank crack-density tensor (scalar crack model) or by a second- and fourth-rank crack-density tensor (joint inversion model). Then the inversion results are used to explore and predict the stress-dependent elastic behavior of various sedimentary rock lithologies using an analytical microstructural rock-physics model via the initial modelinput parameters: initial crack aspect ratio and initial crack density. Estimates of initial crack aspect ratio are consistent among most lithologies with a mean of 0.0004, but for shales they differ up to several times in magnitude with a mean of 0.001. Estimates of initial aspect ratio are relatively insensitive to the inversion method, although the scalar crack inversion becomes less reliable at low values of normal-to-tangential crack compliance ratio (BN/BT) . Initial crack density is sensitive to the degree of damage as well as the inversion procedure. An important implication is that the fourth-rank crack-density term is not necessarily negligible for most sedimentary rocks and evaluation of this term or BN/BT is necessary for accurate prediction of initial crack density. This is especially important because recent studies suggest that BN/BT can indicate fluid content in cracks.