Recently, rare-earth elements (REEs) have attracted great interest due to their importance in several fields, such as the high-technology and medicine industries. Due to the recent intensification of the use of REEs in the world and the resulting potential impact on the environment, new analytical approaches for their determination, fractionation and speciation are needed. Diffusive gradients in thin films are a passive technique already used for sampling labile REEs, providing in situ analyte concentration, fractionation and, consequently, remarkable information on REE geochemistry. However, data based on DGT measurements until now have been based exclusively on the use of a single binding phase (Chelex-100, immobilized in APA gel). The present work proposes a new method for the determination of rare earth elements using an inductively coupled plasma‒mass spectrometry technique and a diffusive gradients in thin films (DGT) technique for application in aquatic environments. New binding gels were tested for DGT using carminic acid as the binding agent. It was concluded that acid dispersion directly in agarose gel presented the best performance, offering a simpler, faster, and greener method for measuring labile REEs compared to the existing DGT binding phase. Deployment curves obtained by immersion tests in the laboratory show that 13 REEs had linearity in their retention by the developed binding agent (retention x time), confirming the main premise of the DGT technique obeying the first Fick's diffusion law. For the first time, the diffusion coefficients were obtained in agarose gels (diffusion medium) and carminic acid immobilized in agarose as the binding phase for La, Ce, Pr, Nd, Sm, Eu, Gd, Dy, Ho, Er, Tm, Yb and Lu, which were 3.94 × 10−6, 3.87 × 10−6, 3.90 × 10−6, 3.79 × 10−6, 3.71 × 10−6, 4.13 × 10−6, 3.75 × 10−6, 3.94 × 10−6, 3.45 × 10−6, 3.97 × 10−6, 3.25 × 10−6, 4.06 × 10−6, and 3.50 × 10−6 cm2 s−1, respectively. Furthermore, the proposed DGT devices were tested in solutions with different pH values (3.5, 5.0, 6.5 and 8) and ionic strengths (I = 0.005 mol L−1, 0.01 mol L−1, 0.05 mol L−1 and 0.1 mol L−1 – NaNO3). The results of these studies showed an average variation in the analyte retention for all elements at a maximum of approximately 20% in the pH tests. This variation is considerably lower than those previously reported when using Chelex resin as a binding agent, particularly for lower pH values. For the ionic strength, the maximum average variation was approximately 20% for all elements (except for I = 0.005 mol L−1). These results indicate the possibility of a wide range of the proposed approach to be used for in situ deployment without the use of correction based on apparent diffusion coefficients (as required for using the conventional approach). In laboratory deployments using acid mine drainage water samples (treated and untreated), it was shown that the proposed approach presents excellent accuracy compared with data obtained from Chelex resin as a binding agent.