This study is a case study of federal logistics support during Hurricane Katrina disaster relief operations. Data from federal contracts covering the first ten weeks of Katrina are used to measure federal logistics activity. The study investigates whether chaos theory, part of complexity science, can extract information from Katrina contracting data to help managers make better logistics decisions during disaster relief. The study uses three analytical techniques: embedding, fitting the data to a logistic equation, and plotting the limit-cycle. Embedding and fitting a logistic equation to the data were used to test for deterministic chaos. The logistic equation and two versions of the limit-cycle model developed by Priesmeyer, Baik and Cole were also tested as potential management tools. This study found deterministic chaos was present during the first week of disaster relief, but inconclusive results for subsequent weeks possibly due to internal changes to the relief dynamics. The research concludes that the initial conditions and early actions will have a significant affect on disaster relief outcome. Furthermore, many events that appear to be uncontrollable and random may actually be controllable.