Kinesin-1 is a molecular transporter that trafficks along microtubules. There is some evidence that kinesin-1 targets specific cellular sites, but it is unclear how this spatial regulation is achieved. To investigate this process, we used a combination of in vivo imaging of kinesin heavy-chain Kif5c (an isoform of kinesin-1) fused to GFP, in vitro analyses and mathematical modelling. GFP-Kif5c fluorescent puncta localised to a subset of microtubules in live cells. These puncta moved at speeds of up to 1 microm second(-1) and exchanged into cortically labelled clusters at microtubule ends. This behaviour depended on the presence of a functional motor domain, because a rigor-mutant GFP-Kif5c bound to microtubules but did not move along them. Further analysis indicated that the microtubule subset decorated by GFP-Kif5c was highly stable and primarily composed of detyrosinated tubulin. In vitro motility assays showed that the motor domain of Kif5c moved detyrosinated microtubules at significantly lower velocities than tyrosinated (unmodified) microtubules. Mathematical modelling predicted that a small increase in detyrosination would bias kinesin-1 occupancy towards detyrosinated microtubules. These data suggest that kinesin-1 preferentially binds to and trafficks on detyrosinated microtubules in vivo, providing a potential basis for the spatial targeting of kinesin-1-based cargo transport.