Little is known about ecosystem structure and nutrient flux in the pelagic zone of seasonal upwelling systems in the tropics, despite their global importance to marine production. The Tropical Eastern Pacific (TEP) is responsible for around 10% of global ocean productivity, largely due to wind-driven seasonal upwelling areas between Mexico and Panama. The Gulf of Panama has a detectable outflow for hundreds of kilometres into the Pacific Ocean and the upwelling system there is an important productivity source in the TEP. In this study, we aim to determine the spatio-temporal patterns in variability of carbon and nitrogen stable isotope composition throughout the pelagic ecosystem of the Bay of Panama in upwelling and non-upwelling conditions, and how these patterns are recorded throughout the ecosystem from primary producers to apex predators. We characterise the stable isotope composition of basal production in the ecosystem to quantify spatial variability during the non-upwelling season. We use the delta N-15 composition outside of upwelling season as an ecosystem baseline and quantify the overall delta N-15 separation between all trophic levels (TL) from primary producers to apex predators (e.g. yellowfin tuna and mahi mahi) within the pelagic ecosystem (2.9 +/- 0.1 parts per thousand per TL). The ecosystem has a relatively simple, linear structure with size-based TL increase. We calculate the predator-prey mass ratio of this ecosystem (c. 113:1 for fishes, 376:1 for the whole ecosystem, uncertainty range: 77:1-1272:1). These values are low for existing estimates in other marine ecosystems, although within the expected range for animals of the mass sampled. The calculated predator-prey mass ratios and maximum TL indicate that this pelagic ecosystem may have a relatively long trophic chain, with inefficient nutrient transfer from low to high TLs. Using a monthly time series of stable isotope values of resident, planktivorous fishes and co-occurring sea surface temperature measurements, we determine nutrient transfer time between primary production and TL3 fish. We calculate a rapid nutrient turnover time within this ecosystem as 0.5-1 month per TL. The incorporation of upwelled nutrients, which are enriched in the C-13, leads to higher consumer delta C-13 values. Carbon isotopes can therefore be used to track timing, duration, and use of upwelling zones. Our findings give novel, empirical insights into the functional ecology of the pelagic ecosystem in the Gulf of Panama, and provide a baseline for comparison and quantification of ecosystem structure and dynamics in the tropics, and in other pelagic upwelling systems.