We study the spectral-decomposition response to reservoir fluids from a deepwater West Africa reservoir through a systematic modeling approach. Our workflow starts from selecting the seismic data (far-angle seismic images) that show more pronounced fluid effect based on amplitude-versus-offset (AVO) analysis. Synthetic seismic forward modeling performed at the control well established the quality of the seismic well tie. Reservoir wedge modeling, spectral decomposition of the field and synthetic seismic data, and theoretical analyses were conducted to understand the spectral-decomposition responses. The reservoir fluid type is a main factor controlling the spectral response. For this deepwater reservoir, the amplitude contrast between oil sand and brine sand is higher at low frequencies . In addition, synthetic modeling can help identify the possible frequency band where the amplitude contrast between hydrocarbon sand and brine sand is higher. When properly included in a comprehensive direct-hydrocarbon-indicator (DHI)–AVO evaluation, spectral decomposition can enhance the identification of hydrocarbons.