Floodplain forests and the woody debris they produce are major components of riverine ecosystems in many arid and semiarid regions (drylands). We monitored breakdown and nitrogen dynamics in wood and bark from a native riparian tree, Fremont cottonwood (Populus deltoides subsp. wislizeni), along four North American desert streams. We placed locally-obtained, fresh, coarse material [disks or cylinders (∼500–2000 cm³)] along two cold-desert and two warm-desert rivers in the Colorado River Basin. Material was placed in both floodplain and aquatic environments, and left in situ for up to 12 years. We tested the hypothesis that breakdown would be fastest in relatively warm and moist aerobic environments by comparing the time required for 50% loss of initial ash-free dry matter (T₅₀) calculated using exponential decay models incorporating a lag term. In cold-desert sites (Green and Yampa rivers, Colorado), disks of wood with bark attached exposed for up to 12 years in locations rarely inundated lost mass at a slower rate (T₅₀ = 34 yr) than in locations inundated during most spring floods (T₅₀ = 12 yr). At the latter locations, bark alone loss mass at a rate initially similar to whole disks (T₅₀ = 13 yr), but which subsequently slowed. In warm-desert sites monitored for 3 years, cylinders of wood with bark removed lost mass very slowly (T₅₀ = 60 yr) at a location never inundated (Bill Williams River, Arizona), whereas decay rate varied among aquatic locations (T₅₀ = 20 yr in Bill Williams River; T₅₀ = 3 yr in Las Vegas Wash, an effluent-dominated stream warmed by treated wastewater inflows). Invertebrates had a minor role in wood breakdown except at in-stream locations in Las Vegas Wash. The presence and form of change in nitrogen content during exposure varied among riverine environments. Our results suggest woody debris breakdown in desert riverine ecosystems is primarily a microbial process with rates determined by landscape position, local weather, and especially the regional climate through its effect on the flow regime. The increased warmth and aridity expected to accompany climate change in the North American southwest will likely retard the already slow wood decay process on naturally functioning desert river floodplains. Our results have implications for designing environmental flows to manage floodplain forest wood budgets, carbon storage, and nutrient cycling along regulated dryland rivers.