Data broadcast is an effective approach to disseminate information from a database server to numerous mobile clients in a mobile environment. Since a broadcast session contains only a subset of the database items, a client might not be able to obtain all its items from the broadcast and is forced to request additional ones from the server on demand.In this paper, we describe a semantic-based broadcast approach which attaches a semantic description to each broadcast unit, called chunk, which is a cluster of data items, This allows a client to determine if a query can be answered entirely using a broadcast as well as defining the precise nature of the remaining items in the form of a “supplementary” query. Chunks could be of different sizes and are hierarchically organized.We propose a heuristic to schedule the broadcast order of the chunks to improve the tuning time, access time, and a new metric called data affinity index. The performance is evaluated via experiments based on a simulation model.