R-OLL (short for "Recognition before OLL") is a recognition technique for the Fridrich Method that allows the solver to anticipate the PLL corner permutation type before OLL. It was proposed by Sébastien Felix in 2007.
R-OLL depends on knowing the effect on corner permutation of the OLL algorithms used. The solver recognizes the corner permutation before OLL (as in CxLL) to determine the resulting corner permutation after OLL. Sébastien Felix organized the then-mainstream OLL algorithms into groups depending on their effect on corner permutation. With practice, R-OLL takes no longer than usual OLL recognition. Together with the French PLL recognition, which always places solved adjacent corners on the left, R-OLL allows one to determine the AUF before PLL recognition.
Very few fast cubers have learned to use R-OLL, Edouard Chambon being one of the few notables. Some top cubers since 2008 have instead opted to learn OLLCP (usually in part), in which multiple OLL algorithms are used to control or solve the corner permutation.