A Raman-pump frequency modulation scheme and an automated signal-processing algorithm are developed for improved collection of time-resolved femtosecond stimulated Raman spectra. Together, these two advancements remove the broad background signals endemic to FSRS measurements and retrieve,signals with high sensitivity. We apply this frequency-modulated femtosecond stimulated Raman spectroscopy (FM-FSRS) to the characterization of ultrafast energy transport in a copolymer comprised of polystyrene linked oligo(phenylene-ethynylene) donor and thio Phene benzothiadiazole acceptor chromophores. After photoexcitation of the donor, ultrafast energy transfer is monitored by the decay of donor vibrational modes and simultaneous growth of acceptor modes. The FM-FSRS method shows clear advantages in signal-to-noise noise levels, mitigation of artifact features, and ease of data processing over -ttheconventional FSRS technique.