In 1940, C.S. Hanes found that potato phosphorylase transferred a few two to three glucose units from α‐D‐Glc‐1‐P to the nonreducing‐ends of starch. Starting with only α‐D‐Glc‐1‐P and phosphorylase, there was no reaction and a starch‐primer was required for reaction. In 1960, Leloir and coworkers incubated starch granules with ADP‐[14C]Glc and found that 14C‐starch was synthesized. Reaction with β‐amylase gave 14C‐maltose from the nonreducing‐ends and it was assumed that the glucose was being added to the nonreducing‐ends of a primer. Mukerjea and Robyt pulsed and chased starch‐granules with ADP‐[14C]Glc and ADPGlc, and found on reduction and hydrolysis of the 14C‐starch, 14C‐D‐glucitol was obtained, indicating that glucose was added to the reducing‐ends of growing starch‐chains. They obtained highly purified, starch‐synthase that was shown to be free of starch‐primers and gave de novo synthesis, with the addition of glucose to the reducing‐ends. They also found 61 papers from 1964 to 2012 that used 50–100 mM Tris‐type buffers with starch‐synthase and had to add putative‐primers to obtain activity. Mukerjea et al. showed that 25 mM Tris‐type buffers, completely inhibited starch‐synthase. Addition of 10 mg/mL putative‐primers, glycogen or maltotetraose, gave partial (8–57%) reversal of the inhibition. The putative‐primers were activators and not primers; however, their use perpetuated the primer myth for 50 years. A hypothesis is developed, as to how starch granules are initiated and grow in vivo.