Biological nitrogen (N) fixation of atmospheric N2 by free-living cyanobacteria in aquatic environments is common, and in many ecosystems, it can account for a significant portion of the biologically available N inputs. Although N fixation can compensate for N limitation, N limitation is maintained over relatively long time scales in many oligotrophic lake ecosystems. This dissertation examines the importance of benthic and pelagic N fixation in the N economies of oligotrophic lakes in arctic Alaska (Chapter 1) and examines nutrient, light, and grazer controls on benthic N fixation (Chapters 2 and 3). Both benthic and pelagic N fixation are prevalent in many lakes across the Alaskan arctic landscape, ranging from 0.12 ? 1.5 mg N m-2 day-1 and 0 ? 2.56 mg N m-2 day-1 respectively. Pelagic N fixation is much higher than has been reported elsewhere for oligotrophic lakes, and is more important than previously thought, comprising ~ 75% of N inputs to one lake. Benthic N fixation is lower than has been reported for other oligotrophic systems, and is roughly equivalent to N inputs from atmospheric deposition on an areal basis (~25 mg N m-2 year-1). On the landscape scale, N fixation in lakes roughly equal that in terrestrial ecosystems in this Arctic region. Benthic N fixation generally appears to have a saturating response to light availability within individual lakes, but light does not explain variation in benthic N fixation across lakes or years. Whole-lake fertilization and laboratory experiments indicate that P input stimulates benthic N fixation while N input suppresses N fixation when N is added either alone or in conjunction with P in Redfield proportion. Snails at ambient density cause a small decline in benthic N fixation (0.85 ? 1.8% reduction over the summer). These patterns are corroborated in the landscape: lakes on younger surfaces have higher P, more snails, and higher rates of N-fixation than lakes on older surfaces.