This paper develops a uniformly valid inference framework forAlmost Stochastic Dominance (ASD). ASD relaxes classical stochasticdominance by allowing small violations of dominance inequalities, and hasbeen widely used in empirical finance and welfare analysis. However,statistical inference for ASD is challenging because the dominanceconditions involve nonlinear inequality restrictions with boundaryconstraints, leading to nonregular asymptotic behavior. Standard pointwiseasymptotic approximations based on the functional delta method fail todeliver uniform validity, particularly near the boundary of the dominanceregion. We propose a joint testing procedure for the necessary andsufficient conditions characterizing ASD and establish its uniformasymptotic validity under general sampling schemes, including weaklydependent time series. Our approach employs uniform bounding and bootstrapmethods that remain valid under drifting sequences of distributions and donot rely on least-favorable configurations alone. We further developuniformly valid inference for a \textit{measure of deviation} from ASD,which quantifies the minimal tolerance level required for almost dominanceto hold. The framework is extended to introduce inference for an \textit{ASDhorizon index}, which identifies the minimal investment horizon at which onedistribution almost stochastically dominates another. Monte Carlosimulations demonstrate that the proposed procedures achieve accurate sizecontrol and improved power relative to existing methods. An empiricalapplication illustrates the relevance of uniform inference forhorizon-dependent dominance relations.