Fisheries typically experience large changes over time in fishing effort. The size of the area fished may also change substantially over time, mimicking the trend in fishing effort, and may have major effects on the population dynamics and fishing process. We extend a biomass dynamics model to incorporate fished area and relative fish density in fished and unfished areas. The fishable population is defined as those individuals in the fished area and those that are sufficiently close to the fished area that they could potentially move into fished area during the fishing season. We estimate fishable biomass using three models assuming different level of population mixing between fished and unfished areas (i.e., partial mixing, full mixing, and no mixing). The models are implemented within a hierarchical Bayesian framework. Model performance is explored using simulations, and the approach is illustrated using logbook data for two tiger prawn species in Australia’s Northern Prawn Fishery. The partial mixing model that involves estimating a mixing parameter performs better than the models that assume no or full mixing. The methods could be applied to other fisheries where the area fished has changed substantially over the history of the fishery.