Zebrafish epiboly involves the vegetal movement of the blastoderm and the yolk syncytial layer to enclose the yolk cell. An elaborate longitudinal microtubule array extends from the yolk syncytial layer towards the vegetal pole of the yolk cytoplasmic layer and shortening of this array is proposed to provide a vegetally directed pulling force during epiboly. Despite experimental evidence implicating yolk microtubules in normal epiboly movements, their exact function remains unclear. Furthermore, most studies have examined yolk microtubules in fixed specimens, thus little is known about their dynamics. To investigate yolk cell microtubule function, we examined the movements of the microtubule plus end associated protein EB3 fused to GFP (EB3-GFP), which binds to actively polymerizing microtubules. The current model is that the microtubule array is established before epiboly and shortens during epiboly and thus predicts very little polymerization during early epiboly stages. In contrast to this model, EB3-GFP tracking revealed widespread microtubule polymerization in the yolk syncytial layer and extensive vegetally directed microtubule growth in the yolk cytoplasmic layer from high stage to 60% epiboly. Strikingly, after 60% epiboly, EB3-GFP was not detected in the yolk cytoplasmic layer, suggesting that the yolk microtubules might be stabilized at 60% epiboly. To investigate this further, we performed antibody staining for tyrosinated and de-tyrosinated tubulin, which are markers of dynamic and stabilized microtubules, respectively. The marker for dynamic microtubules was present at early but not late epiboly stages, while the marker for stabilized microtubules was detected at late but not early epiboly stages. We also performed FRAP experiments on embryos with fluorescently labeled microtubules and found that the recovery times were faster at early epiboly stages than at late epiboly stages. Our data are consistent with the hypothesis that yolk cell microtubules are more dynamic at early epiboly stages than at late epiboly stages, suggesting that they may perform distinct functions at these stages. Previous work suggested that the yolk cell microtubules might pull the yolk syncytial layer nuclei towards the vegetal pole starting at 60% epiboly. Thus, the yolk cell microtubules might need to be stabilized before they can function in yolk syncytial layer nuclear migration, a possibility that we are currently investigating. Overall, we found that the yolk microtubules appear to undergo a rapid growth phase during early epiboly followed by a more stabilized phase that correlates with downward movement of the yolk syncytial layer nuclei. Intriguingly, the two phases of microtubule dynamics are consistent with the timing of the initiation and progression phases of epiboly.