In spoiled gradient echo sequences, the T-1-weighting of image contrast is strongly affected by a nonlinear interaction of two sequence parameters, repetition time (TR) and flip angle (a). If a is not properly adjusted to compensate for variation in TR due to changing resolution, bandwidth, or number of slices, any optimization of contrast-to-noise may be compromised. Currently, there is no direct way to compare or reproduce the contrast properties of one sequence to another with a different TR. Here, it is demonstrated that for short TRs a may be calculated and automatically adjusted such that relative contrast-the shape of the signal-versus-T-1 curve-remains consistent and signal scales proportionally to root TR in all tissues. TR is then free to vary to accommodate a range of sequence parameters without impacting relative contrast and the T-1-weighting of one sequence can be compared to, or reproduced in, another study with a different TR. (c) 2005 Wiley-Liss, Inc.