Using synthetic polypeptides as a model system, the theories of helix-coil transition were developed into one of the most beautiful and fruitful subjects in macromolecular science. The classic models proposed by Schellman and Zimm-Bragg more than 50 years ago, differ in the assumption on whether the configuration of multiple helical sequences separated by random coil sections is allowed in a longer polypeptide chain. Zimm also calculated the critical chain lengths that facilitate such interrupted helices in different solvent conditions. The experimental validation of Zimm's prediction, however, was not carefully examined at that time. Herein, we synthesize a series of homopolypeptide samples with different lengths, to systematically examine their helix-coil transition and folding cooperativity in solution. We find that for longer chains, polypeptides do exist as interrupted helices with scattered coil sections even in helicogenic solvent conditions, as predicted in the Zimm-Bragg model. The critical chain lengths that facilitate such interrupted helices, however, are substantially smaller than Zimm's original estimation. The inaccuracy is in part due to an approximation that Zimm made in simplifying the calculation. But more importantly, we find there exist intramolecular interactions between different structural segments in the longer polypeptides, which are not considered in the classic helix-coil theories. As such, even the Zimm-Bragg model in its exact form cannot fully describe the transition behavior and folding cooperativity of longer polypeptides. The results suggest that long "all-helix" chains may be much less prevalent in solution than previously imagined, and a revised theory is required to accurately account for the helix-coil transition of the longer chains with potential "non-local" intramolecular interactions.