We review our joint experimental-theoretical effort on the folding of photo-switchable α-helices. The folding kinetics of these peptides is profoundly non-exponential, which is attributed to a partitioning of the unfolded state into several misfolded traps. These traps are connected to the folded state in a hub-like fashion with folding barriers of different heights. Molecular dynamics simulations reveal a semi-quantitative agreement with the complex response observed in the experiment, allowing one to discuss the process in unprecedented detail. It is found that the nonexponential response is to a large extent introduced by the photo-linker used to initiate folding. Hence, folding of these cross-linked peptides emulates formation of a helical segment in the context of a globular protein rather than folding of an isolated peptide.