Desorption experiments performed on four Cu-adsorbed palygorskites suggest that the leached Cu2+ ion originates at the surface and/or net-like interstice of the palygorskite fibres. The leached fraction, calculated from the quantities of adsorbed Cu2+ before and after desorption, is <1%. This may indicate that the majority of Cu is in inaccessible structural sites. X-ray diffraction, X-ray photoelectron spectroscopy, Fourier transform infrared (FTIR) spectroscopy and electron spin resonance (ESR) were used to determine the mineralogical character of the Cu-adsorbed palygorskite. Two photoelectron lines at 932.5 and/or 933.7 eV in the narrow scan Cu 2p3/2 spectra show that Cu adsorbed on the surface of palygorskite is in the Cu+ and Cu2+ state. The stretching vibrations of the octahedral cation shift ~3–5 cm−1 towards a greater wavenumber in the FTIR spectra of Cu-adsorbed palygorskite. It can be deduced that the Cu2+ is trapped in the channel of the palygorskite structure. The ESR spectra of the palygorskite give g values of 2.34, 2.12, 2.08 and 2.05, suggesting that some Cu ions cannot be reached by H+. These results confirm that Cu is adsorbed by palygorskite via three possible mechanisms: (1) the Cu is adsorbed onto the surface or in a net-like interstice, and its oxidation states are +1 and +2; (2) Cu forms a complex ion −[Cu(H2O)4]2+ or [Cu(H2O)6]2+, and is trapped in the channel; or (3) Cu enters into the hexagonal channel of the tetrahedral sites or the unoccupied octahedral sites of palygorskite.