KIF1A, a kinesin-related motor protein that transports pre-synaptic vesicles in neurons, was originally presumed to translocate along microtubules (MT) as a monomer. Protein structure predictions from its amino acid sequence failed to identify the long coiled-coil domains typical of kinesins, which led researchers to believe it does not oligomerize into the canonical kinesin dimer. However, mounting evidence using recombinant chimeric protein indicates that KIF1A, like conventional kinesin, requires dimerization for fast, unidirectional processive movement along MTs. Because these studies are somewhat indirect, we wished to test the oligomerization state of native KIF1A, and to compare that to full-length recombinant protein. We have performed hydrodynamic analyses to determine the molecular weights of the respective complexes. Our results indicate that most native KIF1A is soluble and indeed monomeric, but recombinant KIF1A is a dimer. MT-binding studies also showed that native KIF1A did not bind to MTs in either the presence of AMP-PNP, apyrase, or adenosine triphosphate (ATP), but recombinant KIF1A bound to MTs most stably in the presence of ATP, indicating very different motor functional states. To further characterize KIF1A's dimerization potential, we prepared peptides corresponding to the neck domains of MmKIF1A and CeUnc104, and by circular dichroism spectroscopy compared these peptides for their ability to form coiled-coils. Interestingly, both MmKIF1A and CeUnc104 neck peptides formed homodimeric coiled-coils, with the MmKIF1A neck coiled-coil exhibiting the greater stability. Collectively, from our data and from previous studies, we predict that native KIF1A can exist as both an inactive monomer and an active homodimer formed in part through its neck coiled-coil domain.