•Peak fine woody fuel loadings occurred 17–18year post-fire in unmanipulated stands.•Salvage logging increased fine fuel loadings 160–237% for 18–22years post-fire.•Maximum 1000-h fuel loadings occurred 24–31years post-fire in unmanipulated stands.•Decomposition reduced loadings by 35–50% of initial snag necromass by peak years.•Salvage logging significantly reduced 1000-h fuel loadings after 7years post-fire. Salvage logging has been proposed to reduce post-fire hazardous fuels and mitigate re-burn effects, but debate remains about its effectiveness when considering fuel loadings are dynamic, and re-burn occurrence is stochastic, in time. Therefore, evaluating salvage loggings capacity to reduce hazardous fuels requires estimating fuel loadings in unmanipulated and salvaged stands over long time periods. We sampled for snag dynamics, decomposition rates, and fuel loadings within unmanipulated high-severity portions of 7 fires, spanning a 24-year chronosequence, in dry-mixed conifer forests of Oregon’s eastern Cascades. We used these estimates to program an empirical model predicting temporal dynamics of fine and coarse woody fuels from sources directly targeted by salvage logging. We simulated 1000 unmanipulated and salvage logging scenarios, with variable snag dynamic rates, for three sample plots spanning a pre-fire biomass gradient. Total surface fine woody fuel loadings peaked 17–18years post-fire (6.76–9.92Mgha−1) in unmanipulated stands; thereafter decay losses exceeded input rates and loadings decreased. Salvage logging immediately increased surface fine woody fuel loadings by 160–237% above maximum loadings observed in unmanipulated stands, and were higher during the initial 18–22years post-fire. 1000-h fuel loadings peaked 24–31years post-fire in unmanipulated stands, but decomposition reduced total loadings by 34.8–49.6% of initial snag necromass by peak years. Our simulations suggest only 59.2% (34.3Mgha−1), 47.8% (44.3Mgha−1) and 22.3% (37.6Mgha−1) of maximum 1000-h fuel loadings were available for combustion at this time. Available 1000-h fuel loadings in unmanipulated stands peaked 31, 34 and 82years post-fire but were only 35.4% (40.7Mgha−1), 30.8% (49.9Mgha−1) and 27.3% (70.5Mgha−1) of initial snag necromass. Salvage logging increased 1000-h fuel loadings for the initial 7years post-fire, but 80–84% of initial snag necromass was removed or decayed when their maximum loadings were observed 17–22years post-fire. Understory woody vegetation reestablished quickly following high-severity fire, creating another significant fuel layer and a source of post-fire fine woody fuels. Surface fuels accumulate quickly following high-severity fire, but our modeling results suggest salvage logging has mixed effects on reducing hazardous fuel conditions since it increases fine woody fuel loadings and decreases coarse woody fuel loadings. Reducing hazardous fuel loadings and their contribution to re-burn hazard requires manipulation of residual and future fuel sources, but treatment benefits should be evaluated against any negative effects to early seral forest structure and function if resilient forest ecosystems are the management goal.