The polymerization of tubulin has been studied with a high-sensitivity differential scannin microcalorimeter, with results which indicate that microtu ule assembly can proceed via one or possibly two exothermic reactions. The amount of heat evolution has been found to be far in excess of GTP hydrolysis. The heat liberated has been observed to depend strongly upon the exact experimental conditions, varying from many hundreds of kilocalories per mole of tubulin dimer when dilute tubulin solutions are heated rapidly to a few kilocalories per mole of tubulin dimer when concentrated tubulin solutions are heated slowly. The results are tentatively interpreted in terms of the existence of at least two pathways or the ormation of energetically distinct polymers. These findings indicate the importance of kinetic factors in studying tubulin polymerization.Microtubules are a class of organelles that appear to be present in all eukaryotic cells. These structures are thought to function as a cytoskeleton (to maintain the shape and form of cells) or in the transport of various macromolecules and supramolecular structures [for a recent review of these organelles see Dustin (1)]. Because the temporal and spatial control of the assembly and disassembly of microtubules is believed to be a key factor in cellular function, an understanding of the mechanism or mechanisms involved in this process is of great importance.Since the discovery of the conditions necessary for the in vitro assembly of cytoplasmic microtubules (2), a considerable amount of biochemical and biophysical work has been directed toward unraveling the mechanism of the polymerization and depolymerization of tubulin, the major protein subunit of microtubules, which has a molecular weight of 110,000 (3). Although various models have been proposed [which are summarized in a recent review by Kirschner (4)], there is no general agreement as to their validity. In all these models it has been assumed that the assembly of microtubules is reversible, and that its temperature-dependent behavior shows it to be an endothermic reaction. However, if GTP hydrolysis is indeed coupled to this process, a question arises as to its thermodynamic reversibility. The recent results of Margolis and Wilson (5), which showed that the pathway for microtubule assembly is not the pathway for microtubule disassembly, raise further questions concerning reversibility.