Count data models that incorporate the existence of excess zero observations were employed to estimate minimum demand thresholds for rural Wisconsin retail sectors. It was found that single- and double-hurdle models improved the estimation of rural retail minimum demand thresholds.