Spoiled fast low‐angle shot (FLASH) magnetic resonance imaging (MRI) provides a simple contrast, largely independent of the spin‐spin relaxation time. Three‐dimensional FLASH imaging (3D FLASH) has gained importance due to its simplicity and its ability to produce isotropic, high‐resolution images. Nevertheless, 2D FLASH imaging can be more suited than 3D FLASH imaging in various specific applications requiring long repetition time, such as imaging of the proton density (PD), as it allows interleaved slice acquisition rather than consecutive slice acquisition. In practice, however, a slice‐selective excitation produces a nonuniform excitation profile, which needs to be taken into account to allow reliable quantitative data analysis. In this work, the influence of the nonuniform excitation profile on the detected signal is modeled as a dimensionless multiplicative correction, function of the flip angle, and the ratio of the repetition time to the spin‐lattice relaxation time T1. This model is validated experimentally by measuring the PD and T1 MR parameters in a phantom experiment on a 3T clinical scanner. A good accuracy in the estimation of T1 and in the reconstruction of the PD (weighted by the receiver sensitivity profile) is obtained, for example, a relative error of 3% in T1 within the range [500, 2000] ms and a precision of 1% in PD. It is also shown how the proposed theory can be extended to magnetization‐prepared 2D‐spoiled FLASH as well as other variations of the 2D‐spoiled FLASH sequence. © 2013 Wiley Periodicals, Inc. Concepts Magn Reson Part A 42A: 89–100, 2013.