Stable isotope ratios of water (𝛿 18O,𝛿 2H) have long been used to study a core question in plant ecology and ecohydrology: “From where do plants take up water?” Indeed, decades of research has involved sampling potential plant water sources in the subsurface, classifying those sources as distinct endmembers (e.g., deep versus shallow soil waters), and then evaluating their contributions to a xylem water sample through mixing‐model analysis to identify the depths of root water uptake. However, more detailed interrogations of the subsurface and plant domains have revealed under‐considered transport and isotopic‐fractionation phenomena. These now apparent complexities raise new questions and challenge the many past assumptions inherent in endmember‐mixing models that now seem overly simple. Here, we introduce discussions of these recent insights and provide an overview of isotope effects that occur naturally in the root zone and in the plant, as well as artificially during sample handling. Better accounting for these complexities and their associated uncertainties can lead to more accurate and robust study designs, analytical frameworks, and, ultimately, inferences. Finally, to more robustly characterize plant water sources using 𝛿 18O and 𝛿 2H, we provide some practical recommendations that aim at maximizing the isotopic contrast between endmembers and/or minimizing potential uncertainties.