This paper consistently estimates key structural properties of a multiple-species fishing technology. We overcome two ubiquitous features of fisheries data generating processes that invalidate classical estimation of fishing technologies: unobservability by the researcher but partial observability of the fish stock by fishermen and endogenous production decisions that vary with fishermen’s private knowledge of true stock abundance. Our identification strategy exploits timing and available information when production decision are made, technological constraints, and natural, exogenous variability of fish stock abundance. Consistency in estimation obtains under reasonable assumptions for fisheries data generating processes. An application to the U.S. Gulf of Mexico commercial reef fish fishery is presented to demonstrate our approach and reveal substantial bias under estimators that ignore the problem of omitted stock abundance. Implications for improved fisheries management are discussed.