​Polar areas are experiencing the steepest warming rates on Earth, a trend expected to continue in the future. In these habitats, phytoplankton communities constitute the basis of the food web and their thermal tolerance may dictate how warming affects these delicate environments. Here, we compiled available data on thermal responses of phytoplankton growth in polar waters. We assembled 53 growth-vs.-temperature curves (25 from the Arctic, 28 from the Southern oceans), indicating the limited information available for these ecosystems. Half of the data from Arctic phytoplankton came from natural communities where low ambient concentrations could limit growth rates. Phytoplankton from polar waters grew faster under small temperature increases until reaching an optimum (TOPT), and slowed when temperatures increased beyond this value. This left-skewed curves were characterized by higher activation energies (Ea) for phytoplankton growth above than below the TOPT. Combining these thermal responses we obtained a community TOPT of 6.5°C (±0.2) and 5.2°C (±0.1) for Arctic and Southern Ocean phytoplankton communities, respectively. These threshold temperatures were already exceeded at 70°N during the first half of August 2013, evidenced by sea surface temperatures (SSTs, satellite data, http://www.ncdc.noaa.gov). We forecasted SSTs for the end of the twenty-first century by assuming an overall 3°C increase, equivalent to a low emission scenario. Our forecasts show that SSTs at 70°N are expected to exceed TOPT during summer by 2100, and during the first half of August at 75°N. While recent Arctic spring temperatures average 0.5°C and −0.7°C at 70°N and 75°N, respectively, they could increase to 2.8°C at 70°N and 2.2°C at 75°N as we approach 2100. Such temperature increases could lead to intense phytoplankton blooms, shortened by fast nutrient consumption. As SSTs increase, thermal thresholds for phytoplankton growth would be eventually exceeded during bloom development. This could lead to changes in the blooming phytoplankton community, threatening the production peak and cycles in the Arctic. Our forecasted phytoplankton responses, are constrained by the limited data set, besides uncertainties in the most plausible future Arctic temperature scenarios. To improve predictions in polar oceans, we need to increase the number of studies, in particular for a fast-changing Arctic.