<p>Length-based stock assessment methods are widely used in data-limited fisheries, yet their effectiveness under uncertain life-history parameters remains inadequately assessed. Such uncertainty is often unavoidable in these contexts and may bias assessment outcomes or misguide management decisions. This study addresses this gap by evaluating the performance of three prevalent methods—Length-Based Spawning Potential Ratio (LBSPR), Length-Based Bayesian Biomass Estimation (LBB), and Length-Based Risk Analysis (LBRA)—under varying levels of uncertainties in key life-history parameters. Utilizing Monte Carlo simulations with Latin hypercube sampling, we systematically test the accuracy and robustness of each method for short-, medium-, and long-lived species, with additional consideration of the potential violations of equilibrium assumptions in stock dynamics. Generalized linear models were used to examine the influence of uncertainty in growth coefficient (<i>K</i>), asymptotic length (<i>L</i>_<i>inf</i>), and natural mortality rate (<i>M</i>), along with their interaction effects. Our results indicate that LBRA showed the most consistent robustness across species and scenarios. LBSPR performed well for medium- and long-lived species but was sensitive to uncertainty in K. LBB was effective only under equilibrium conditions and showed poor performance for long-lived species. Notably, the performance of all three methods was strongly influenced by interactions between Linf and K, and between M and K, highlighting the importance of accounting for interaction effects among life-history parameters. These findings emphasize the need to select assessment methods that are suited to species-specific traits and patterns of uncertainty, in order to enhance the reliability of length-based assessments in data-limited fisheries.</p>