Accurate extraction of soil and plant water for stable isotope analysis remains a methodological challenge in ecohydrology, particularly due to isotopic biases introduced by heating or selective pore-water extraction in conventional techniques. This study developed and evaluated a cryogenic airtight vapor extraction (CRAVE) method from soil and vegetation samples at ambient temperature within a recirculating vapor-liquid pathway. This approach avoids heating-induced non-equilibrium effects and reduces matrix-dependent artifacts and organic contamination, thereby facilitating direct comparison of isotopic compositions between soil and plant water. The results demonstrate that CRAVE-derived isotopic signatures align with both cryogenic vacuum distillation (CVD) and suction lysimeter (SL) benchmarks. However, systematic deviations were observed based on specific matrix properties. For xylem water, the d<sup>2</sup>H offset between CRAVE and CVD was strongly modulated by gravimetric water content (dry-weight basis), with CVD exhibiting greater hydrogen isotope depletion under low-moisture conditions (< 0.8 g·g<sup>-1</sup>). For soils, the isotopic divergence between CRAVE and CVD was driven primarily by soil texture, with offsets increasing as clay content and depth increased (<em>r</em> = 0.82–0.94), where CVD-extracted bulk water became depleted progressively in both δ<sup>2</sup>H and δ<sup>18</sup>O relative to the mobile-capillary pool captured by SL and CRAVE. The Rayleigh-based framework provides a physically grounded means to reconstruct source-water isotope values from condensate measurements; its potential use for mobile–immobile partitioning should, however, be treated as a future application pending targeted validation. Overall, CRAVE represents a promising ambient-temperature extraction method for tracing water partitioning and source-uptake dynamics within the soil–plant–atmosphere continuum.