AbstractThis paper presents the evaluation of two methods for manually estimating the elastic lateral-torsional buckling (LTB) resistance of general doubly and singly symmetric I-section members having a linear variation in the web depth. The study focuses on unbraced lengths having idealized torsionally simply supported (fork) end conditions. The first method uses cross-sectional properties at the midspan of the unbraced segment along with a moment gradient modifier equation, written in terms of the flange stresses, to develop its strength approximation. This approach is employed in the first edition of the AISC-MBMA Design Guide 25 (DG 25). The second method is an extension of an approach that has been shown to be accurate and efficient for prismatic doubly and singly symmetric I-section members and is recommended in the second edition of DG 25. This approach uses cross-sectional properties at a critical section within the unbraced length, defined as the cross section where the ratio of the moment demand to the base LTB resistance in uniform bending is maximum, along with a moment gradient modifier equation written in terms of the above maximum ratio and the corresponding ratios at the quarter points of the unbraced length. Furthermore, Method 2 employs an additional calibration factor that accounts for an apparent destabilizing effect of the inclination of the shear center with respect to the axis of torsional restraint. The predictions from these methods are evaluated by an extensive parametric study encompassing the full design space for I-section members with linearly-varying web depth, as defined by DG 25. The second method is shown to provide improved accuracy, avoiding undesirable nonconservative errors in certain situations. The calibration factor developed as part of the second method is shown to eliminate undesirable errors in the first method as well, providing streamlined and accurate buckling estimates for linearly tapered member unbraced lengths.