The plastic concentration in terrestrial systems is orders of magnitude higher than that found in marine ecosystems, which has raised global concerns about their potential risk to agricultural sustainability. Previous research on the transport of nanoplastics in soil relied heavily on the qualitative prediction of the mean-field extended Derjaguin–Landau–Verwey–Overbeek theory (XDLVO), but direct and quantitative measurements of the interfacial forces between single nanoplastics and porous media are lacking. In this study, we conducted multiscale investigations ranging from column transport experiments to single particle measurements. The maximum effluent concentration (C/C0) of amino-modified nanoplastics (P S–NH 2) was 0.94, whereas that of the carboxyl-modified nanoplastics (P S–COOH) was only 0.33, indicating P S-NH 2 were more mobile than P S-COOH at different ionic strengths (1–50 mM) and pH values (5–9). This phenomenon was mainly attributed to the homogeneous aggregation of P S-COOH. In addition, the transport of P S-NH 2 in the quartz sand column was inhibited with the increase of ionic strength and pH, and pH was the major factor governing their mobility. The transport of P S-COOH was inhibited with increasing ionic strength and decreasing pH. Hydrophilicity/hydrophobicity-mediated interactions and particle heterogeneity strongly interfered with interfacial forces, leading to the qualitative prediction of XDLVO, contrary to experimental observations. Through the combination of XDLVO and colloidal atomic force microscopy, accurate and quantitative interfacial forces can provide compelling insight into the fate of nanoparticles in the soil environment. [Display omitted] • Ionic strength and pH synergistically inhibited the transport of P S-NH 2. • Ionic strength and pH antagonistically controlled the transport of P S-COOH. • Hydrophilicity and surface heterogeneity interfere with XDLVO predictions. • Combination of C -AFM and XDLVO provide accurate understandings for NPs transport. [ABSTRACT FROM AUTHOR]