School of thought in the protein folding community that differs from the classical model of protein folding. The classical model states that folding of a protein occurs through specific intermediates. In other words, a protein folds through a pathway:

unfolded -> intermediate₁ -> intermediate₂ -> folded

The folding funnel hypothesis states that there is no one specific pathway that a protein folds along. Instead, a protein runs down a multi-dimensional funnel towards the final structure. There are a large number of possible intermediates between the fully unfolded and fully folded protein. There is no physical funnel, of course. Its just another way of conceptualizing a complex reaction coordinate.

schematic of a folding funnel

	(------- entropy -------)

^	\			/ - fully unfolded protein
|	 \		       /
|	  -\		      /
|           |		     -
|	    \		    /
| energy     \		   /
|             \            |      many partially folded 
|             |            /          intermediates
|             \           /
|              --\       |
|                 |      /
v                 \     /
                   \   /
                    \ /
                     X - folded protein

The ruggedness of the funnel surface can represent the presence of kinetic traps to folding, where the protein gets stuck in misfolded intermediates. One can think of a folding protein as a person skiing down a slope. If the slope is smooth and steep, the person is going to reach the bottom quickly. If the slope is shallow, or has a lot of bumps and pits, or both, then it will take longer to get to the bottom.