Small boreal lakes emit large amounts of carbon dioxide (CO2) and methane (CH4) to the atmosphere. Yet emissions of these greenhouse gases are variable in space and time, in part due to variable within-lake CO2 and CH4 concentrations. To determine the extent and the underlying drivers of this variation, we measured lake water CO2 and CH4 concentrations and estimated associated emissions using spatially discrete water samples collected every 2 weeks from a small boreal lake. On select dates, we also collected groundwater samples from the surrounding catchment. On average, groundwater draining a connected peat mire complex had significantly higher CO2 and CH4 concentrations compared to waters draining forest on mineral soils. However, within the lake, only CH4 concentrations nearshore from the mire complex were significantly elevated. We observed little spatial variability in surface water CO2; however, bottom water CO2 in the pelagic zone was significantly higher than bottom waters at nearshore locations. Overall, temperature, precipitation, and thermal stratification explained temporal patterns of CO2 concentration, whereas hydrology (discharge and precipitation) best predicted the variation in CH4 concentration. Consistent with these different controls, the highest CO2 emission was related to lake turnover at the end of August while the highest CH4 emission was associated with precipitation events at the end of June. These results suggest that annual carbon emissions from small boreal lakes are influenced by temporal variation in weather conditions that regulate thermal stratification and trigger hydrologic land–water connections that supply gases from catchment soils to the lake.