Glucagon plays a crucial role in regulating glucose homeostasis; unfortunately, the mechanisms controlling its release are still unclear. Capillary electrophoresis (CE)‐ and fluorescence anisotropy (FA)‐immunoassays (IA) have been used for online measurements of hormone secretion on microfluidic platforms, although their use in glucagon assays is less common. We set out to compare a glucagon‐competitive IA using these two techniques. Theoretical calibration curves were generated for both CE‐ and FA–IA and results indicated that CE‐IA provided higher sensitivity than FA–IA. These results were confirmed in an experiment where both assays showed limits of detection (LOD) of 30 nM, but the CE‐IA had ∼300‐fold larger sensitivity from 0 to 200 nM glucagon. However, in online experiments where reagents were mixed within the device, the sensitivity of the CE‐IA was reduced ∼3‐fold resulting in a higher LOD of 70 nM, whereas the FA–IA remained essentially unchanged. This lowered sensitivity in the online CE‐IA was likely due to poor sampling by electroosmotic flow from the high salt solution necessary in online experiments, whereas pressure‐based sampling used in FA–IA was not affected. We conclude that FA–IA, despite lowered sensitivity, is more suitable for online mixing scenarios due to the ability to use pressure‐driven flow and other practical advantages such as the use of larger channels.