Split-spectrum amplitude-decorrelation angiography (SSADA) is a novel method for improving the signal-to-noise ratio (SNR) of flow detection in optical coherence tomography (OCT). SSADA splits the full OCT spectrum into narrower bands, computes inter-B-scan decorrelation separately for each band, and then averages the results. This approach reduces axial motion noise while preserving transverse flow signals, making it particularly effective for retinal and choroidal imaging. The algorithm was tested on in vivo images of the human macula and optic nerve head, showing significant improvements in SNR and microvascular network connectivity compared to other amplitude-decorrelation methods. The study describes a high-speed swept-source OCT system used to acquire 3D volumetric data for SSADA. The system has an axial resolution of 5.3 μm and an imaging range of 2.9 mm. The system was used to image the retinal, choroidal, and optic nerve head regions of normal human subjects. The SSADA algorithm was applied to the data, and the results were compared with other amplitude-decorrelation methods. The theoretical analysis of SSADA involves splitting the OCT spectrum into different frequency bands to reduce axial motion noise. The decorrelation of OCT signal amplitude between B-scans is used to detect flow. The full-spectrum decorrelation method, pixel-averaging, and split-spectrum methods were compared. The split-spectrum method showed the best performance in visualizing the capillary network and reducing noise. The in vivo testing of SSADA demonstrated its effectiveness in visualizing retinal and choroidal vasculature. The SSADA algorithm provided high contrast between flow pixels and non-flow regions, allowing for the visualization of fine vascular structures. The results showed that SSADA could effectively detect capillary networks and differentiate between flow and non-flow pixels, even in regions with low signal intensity. The study also compared the performance of three amplitude-decorrelation angiography algorithms: full-spectrum, pixel-averaging, and split-spectrum. The split-spectrum method provided the cleanest visualization of the capillary network and the best contrast between flow and non-flow pixels. The pixel-averaging method was second best, while the full-spectrum method showed more disconnected flow pixels, likely due to noise. The results indicate that SSADA is a promising technique for improving the SNR and visualization of microvascular networks in OCT imaging.Split-spectrum amplitude-decorrelation angiography (SSADA) is a novel method for improving the signal-to-noise ratio (SNR) of flow detection in optical coherence tomography (OCT). SSADA splits the full OCT spectrum into narrower bands, computes inter-B-scan decorrelation separately for each band, and then averages the results. This approach reduces axial motion noise while preserving transverse flow signals, making it particularly effective for retinal and choroidal imaging. The algorithm was tested on in vivo images of the human macula and optic nerve head, showing significant improvements in SNR and microvascular network connectivity compared to other amplitude-decorrelation methods. The study describes a high-speed swept-source OCT system used to acquire 3D volumetric data for SSADA. The system has an axial resolution of 5.3 μm and an imaging range of 2.9 mm. The system was used to image the retinal, choroidal, and optic nerve head regions of normal human subjects. The SSADA algorithm was applied to the data, and the results were compared with other amplitude-decorrelation methods. The theoretical analysis of SSADA involves splitting the OCT spectrum into different frequency bands to reduce axial motion noise. The decorrelation of OCT signal amplitude between B-scans is used to detect flow. The full-spectrum decorrelation method, pixel-averaging, and split-spectrum methods were compared. The split-spectrum method showed the best performance in visualizing the capillary network and reducing noise. The in vivo testing of SSADA demonstrated its effectiveness in visualizing retinal and choroidal vasculature. The SSADA algorithm provided high contrast between flow pixels and non-flow regions, allowing for the visualization of fine vascular structures. The results showed that SSADA could effectively detect capillary networks and differentiate between flow and non-flow pixels, even in regions with low signal intensity. The study also compared the performance of three amplitude-decorrelation angiography algorithms: full-spectrum, pixel-averaging, and split-spectrum. The split-spectrum method provided the cleanest visualization of the capillary network and the best contrast between flow and non-flow pixels. The pixel-averaging method was second best, while the full-spectrum method showed more disconnected flow pixels, likely due to noise. The results indicate that SSADA is a promising technique for improving the SNR and visualization of microvascular networks in OCT imaging.