Channel avulsions in fluvial systems play a significant role in determining the architecture of alluvial deposits, yet there is insufficient experimental or field data to satisfactorily identify and constrain the controls on the frequency and spatial distribution of channel avulsions over time. Understanding these controls is beneficial scientifically (improving interpretations of the geological record), environmentally (augmenting our understanding of the geomorphic response of altered fluvial systems) and economically (furthering our ability to predict the location and density of alluvial sand bodies that comprise aquifers and petroleum reservoirs). Previous studies have suggested that subsidence rate and geometry serve as primary factors that govern the timing and location of avulsions, as well as the resultant alluvial architecture. To study the relationship between these variables, we used the Experimental Earthscape Facility at the University of Minnesota to model a braided river system under four different subsidence scenarios. For each of the four stages, sediment supply was matched as closely as possible to the rate of creation of accommodation space. Surface processes were recorded on video tape, the surface topography was systematically measured with a laser scanner, and the basement topography was continuously recorded by the subsidence-control system. The final (cumulative) deposit was sectioned and imaged at 2 cm intervals to record the stratigraphy. Preliminary analyses of the surface processes indicate that a change from a laterally variable subsidence pattern to a hinge-like pattern (with similar subsidence rates) had no effect on the frequency or spatial distribution of avulsions. The deposit architecture also seems to be indifferent to the change in subsidence geometry. This may be due to the significantly different timescales over which channel avulsion and subsidence take place. Avulsion frequency sharply increased with a reduction in subsidence rate, however this may have been induced by a marine transgression during this stage of the experiment. A four-fold reduction in sediment concentration (with the subsidence rate held constant) resulted in a four-fold reduction in the frequency of avulsions, but had little effect on their distribution.