Active transport between nucleus and cytoplasm proceeds through nuclear pore complexes (NPCs) and is mediated largely by shuttling transport receptors that use direct RanGTP binding to coordinate loading and unloading of cargo    . Import receptors such as importin beta or transportin bind their substrates at low RanGTP levels in the cytoplasm and release them upon encountering RanGTP in the nucleus, where a high RanGTP concentration is predicted. This substrate release is, in the case of import by the importin alpha/beta heterodimer, coupled directly to importin beta release from the NPCs. If the importin beta -RanGTP interaction is prevented, import intermediates arrest at the nuclear side of the NPCs  . This arrest makes it difficult to probe directly the Ran and energy requirements of the actual translocation from the cytoplasmic to the nuclear side of the NPC, which immediately precedes substrate release. Here, we have shown that in the case of transportin, dissociation of transportin-substrate complexes is uncoupled from transportin release from NPCs. This allowed us to dissect the requirements of translocation through the NPC, substrate release and transportin recycling. Surprisingly, translocation of transportin-substrate complexes into the nucleus requires neither Ran nor nucleoside triphosphates (NTPs). It is only nuclear RanGTP, not GTP hydrolysis, that is needed for dissociation of transportin-substrate complexes and for re-export of transportin to the cytoplasm. GTP hydrolysis is apparently required only to restore the import competence of the re-exported transportin and, thus, for multiple rounds of transportin-dependent import. In addition, we provide evidence that at least one type of substrate can also complete NPC passage mediated by importin beta independently of Ran and energy.