An empirical equation for the magnetization transfer (MT)FLASH signal is derived by analogy to dual-excitation FLASH, introducing a novel semiquantitative parameter for MT, the percentage saturation imposed by one MT pulse during TR. This parameter is obtained by a linear transformation of the inverse signal, using two reference experiments of proton density and T 1 weighting. The influence of sequence parameters on the MT saturation was studied. An 8.5-min protocol for brain imaging at 3 T was based on nonselective sagittal 3D-FLASH at 1.25 mm isotropic resolution using partial acquisition techniques (TR/TE/ ␣ ؍ 25ms/4.9ms/5°or 11ms/4.9ms/15°for the T 1 reference). A 12.8 ms Gaussian MT pulse was applied 2.2 kHz off-resonance with 540°flip angle. The MT saturation maps showed an excellent contrast in the brain due to clearly separated distributions for white and gray matter and cerebrospinal fluid. Magnetization transfer (MT) is a contrast mechanism in tissue that is based on cross-relaxation or chemical exchange between protons in bulk water and rotationally immobilized macromolecules (1,2). In clinical MRI, MT contrast is invoked by application of a T 2 -selective RF pulse applied prior to slice excitation. The "MT pulse" saturates specifically the macromolecular magnetization, and MT is then observed as a reduction in image intensity. Its normalized value, the MT ratio (MTR), is commonly taken as a measure for the strength of the MT effect, and hence interpreted as a surrogate parameter for macromolecular content or myelination. However, the MTR is not an absolute measure, but depends on the sequence parameters and is influenced by T 1 relaxation and flip angle inhomogeneities.This limitation can be overcome by quantitative Z-spectroscopy measurements to obtain parameter estimates for the binary spin-bath (BSB) model and suitable absorption line shapes (3,4). On clinical MR systems, pulsed MT is combined with spoiled gradient echo sequences (FLASH, fast low angle shot) and suitable adaptations of Henkelman's continuous wave model (5-7). Alternatively, pulsed MT experiments can be approximated by instantaneous events of saturation separating intervals free of irradiation (8). Since brain tissue is characterized by conditions of fast-exchange, this "free" evolution can be described by two exponential time courses, the common T 1 relaxation and the MT. The latter is observed as a reduction of the longitudinal magnetization of free water subsequent to the MT pulse (9). This MT-related saturation increases in time until the exchange equilibrium in the BSB model is restored.The concept of separating T 1 relaxation and MT has been previously applied to progressive partial saturation by repetitive MT pulses (10). Here, it is transferred to the MT-w(eighted) FLASH sequence. A phenomenological signal equation is derived by analogy from a FLASH experiment with two interleaved excitations and recovery times. This equation represents the effects of excitation and relaxation during TR, while any additional saturation du...