The authors present a distributed channel acquisition mechanism in dynamic spectrum access (DSA) networks. A novel graph MAximum SPECTrum packing algorithm (MASPECT) is proposed for a system with N secondary networks. Each secondary network makes use only of the local topology information to resolve contentions during channel access. The proposed algorithm also adapts easily to topological changes. Most approaches to channel acquisition use either centralised graph colouring or distributed approaches that result in under utilisation of spectrum. The authors show that the proposed MASPECT algorithm results in improvements by up to one order of magnitude in the spectrum utilisation for the secondary networks and results in Jain's fairness index of about 0.9. The authors further present a modified probabilistic heuristic, PMASPECT, that improves the termination time of the algorithm by up to two orders of magnitude. The authors also study the impact of the MASPECT algorithm on each individual secondary networks in the system in the presence of primary activity. For each secondary network, the proposed algorithm results in up to two orders of magnitude of reduction in the call blocking dropping probabilities in the presence of primary activity.