Aim To compare the performance of oral fundus fluorescein angiography with a confocal scanning laser ophthalmoscope (SLO) with intravenous fundus fluorescein angiography (IVFFA) with a fundus camera in the assessment of sight-threatening diabetic retinopathy. Patients and methods A total of 25 patients undergoing IVFFA to investigate their diabetic retinopathy were recruited. Participants returned 1 week later and an oral angiogram with the SLO was performed. Six facets of the oral and intravenous angiograms were scored and compared: visualization of the foveal avascular zone (FAZ); branch retinal identification; macular leakage; identification of microaneurysms in areas of macular leakage; peripheral nonperfusion, and leakage from neovascular complexes. Results Compared to IVFFA, the FAZ was unreliably visualized with oral angiography (Kappa 0.1, 95% CI 0-0.3). In contrast, macular leakage (Kappa 0.78, 95%, CI 0.72-0.83), identification of microaneurysms in areas of macular leakage (Kappa 0.78, 95%, CI 0.72-0.83), and neovascular complexes (Kappa 1.0) were reliably seen. Analysis of the visualization of peripheral nonperfusion was complicated by the finding that profuse dye leakage from neovascular complexes obscured the view of the peripheral retina. If the five angiograms in which this occurred were excluded, oral angiography identified 23 of the 24 eyes in which significant nonperfusion was found on IVFFA. Conclusion Oral angiography with the SLO can provide high-quality angiograms that allow judgments to be made about the presence of treatable diabetic maculopathy, proliferative diabetic retinopathy, and peripheral nonperfusion. In the presence of coexisting macular oedema, it proved to be an unreliable technique with which to investigate foveal ischaemia.