A reduced-gravity ocean general circulation model of the tropical Pacific Ocean is used to determine potential improvements to the simulated equatorial Pacific cold tongue region through choices in horizontal resolution and coastline geometry—in particular, for the Galápagos Islands. Four simulations are performed, with identical climatological forcing. Results are compared between model grids with and without the Galápagos Islands, with coarse and fine resolutions. It is found that properly including the Galápagos Islands results in the obstruction of the Equatorial Undercurrent (EUC), which leads to improvements in the simulated spatial structure of the cold tongue, including a basinwide warming of up to 2°C in the east-central Pacific. The obstruction of the EUC is directly related to the improvements east of the Galápagos Islands, and for the basinwide reduction of the tropical cold bias through an equatorial dynamical adjustment. The pattern of SST warming resulting from the inclusion of the Galápagos Islands is similar to that of the known cold biases in ocean models and the current National Oceanic and Atmospheric Administration Climate Forecast System. It is thought that such an improvement would have a considerable impact on the ability of coupled ocean–atmosphere and ocean–ecosystem models to produce realistic clouds, precipitation, surface ocean bioproductivity, and carbon cycling in the tropical Pacific Ocean.