Assessments of the cumulative ecological effects of chronic sublethal discharges from normal (non-spill) offshore petroleum operations have been based on (1) model studies, (2) controlled laboratory and microcosm experiments, (3) studies of spills of crude oil and its refined products and of severe organic pollution, and (4) ecological field studies in areas of intensive petroleum operations. Methods (2) and (3) involve (a) the extrapolation of results obtained from concentrations and rates of exposure that are not representative of chronic low-level exposure, (b) usually a different suite of chemical contaminants, and (c) usually a different ecosystem. The fourth method is faced with determining the ecological effects of very low levels of contamination in the presence of larger natural environmental and biological variations and of other sources or types of contamination. This paper compares predictions based on method (3) with results obtained by using method (4) in the Gulf of Mexico. Primary data are from three studies on cumulative effects and from studies of spills in the Gulf. The response of benthic macroinfaunal communities and populations to sediment hydrocarbon concentrations are the primary basis for comparisons. Some principal environmental variables are `normalized' to facilitate quantitative comparisons.