The development of techniques for supporting incremental maintenance of materialized views has been an active research area for over twenty years. However, although there has been much research on methods and algorithms, there are surprisingly few systematic studies on the performance of different approaches. As a result, understanding of the circumstances in which materialized views are beneficial (or not) can be seen to lag behind research on incremental maintenance techniques. This paper presents the results of an experimental performance analysis carried out in a system that incrementally maintains OQL views in an ODMG compliant object database. The results indicate how the effectiveness of incremental maintenance is affected by issues such as database size, and the complexity and selectivity of views.