Decadal variations of very small amplitude (~0.3C in sea surface temperature) in the tropical Pacific Ocean have been shown to have powerful impacts on global climate. Future projections from different climate models do not agree on how this critical feature will change under the influence of anthropogenic forcing. A number of attempts have been made to resolve this issue by examining trends from the 1880s to the present, a period of rising atmospheric concentrations of greenhouse gases. The most recent concluded that the three major data sets disagreed on the trend in the equatorial gradient of sea surface temperature (SST). Using a corrected version of one of these data sets, and extending the analysis to the seasonal cycle, we show here that all agree that the equatorial SST gradient strengthened from 1880–2005, especially during the boreal fall when the gradient is normally strongest. This result appears to favor a theory for future changes based on ocean dynamics over one based on atmospheric energy constraints. Both theories incorporate the expectation that, based on ENSO theory, the zonal sea level pressure (SLP) gradient in the tropical Pacific is coupled to SST and should strengthen along with the SST gradient. We find, however, that the SLP gradient appears to have weakened over the same time period, though consistent with the SST seasonal trends, it weakens least in boreal fall. Most of the IPCC AR4 models capture prominent features of these trends, but they fail to reproduce observed trends in boreal spring.