Purpose: Aim of the study was to evaluate FLAIR com bined with Turbo-Spin-Echo (Turbo-FLAIR) at 0.5 tesla in comparison to conventional T-1- and T-2-weighted spin-echo images (SE) in MRI of the brain. Material and methods: A Turbo-FLAIR sequence was optimised for 0.5 Tesla (Philips Gyroscan T5-II) that provided seventeen 5 mm sections in 4:21 minutes (T-R = 6075 ms, T-1 = 1600 ms, T-E = 120 ms and a turbofactor = 17). Images were compared with T-1- (T-R = 500 ms, T-E = 15 ms) proton density and T-2-weighted spin-echo (T-R = 2500 ms, T-E = 20/90 ms) studies in 10 healthy volun teers and 30 patients with various brain pathologies. Results: Turbo-FLAIR could effectively eliminate the CSF signal in all studies except ventricular areas with CSF inflow. Contrast-to-noise ratios (C/N) for the contrast between lesions and CSF was superior in Turbo-FLAIR. C/N between lesions and grey matter was significantly higher in Turbo-FLAIR than in proton density (PD)-weighted SE. C/N between lesions and white matter was equal to PD-weighted SE but significantly smaller than on T-2-weighted SE. Visual analysis showed greater lesions conspicuity with Turbo-FLAIR and a higher frequency of detection of cortical and subcortical lesions. Conclusions: Turbo-FLAIR is a reliable and practical technique on 6.5 Tesla, that is more sensitive than SE sequences especially for the detection of cortical, subcortical lesions and lesions surrounded by CSF.