We used two measures to compare the effectiveness of 52 conservation criteria in achieving conservation targets for forest types. The first measure was efficiency. Although widely used, efficiency assumes no loss or reduction of biodiversity features before conservation is implemented. This is invalid in many situations. Often, it is more realistic to assume gradual implementation accompanied by incremental, predictable reduction and loss of biodiversity features. We simulated future landscapes resulting from the annual interplay of loss and conservation of forest types. We then based our second measure, retention, on how well criteria scheduled conservation action to prevent targets being compromised. The simulations partly support predictions about the best criteria for scheduling implementation with continuing biodiversity loss. Retention was weakly related or unrelated to efficiency across 52 criteria. Although retention values were sensitive to changes in targets and rates of conservation and forest loss, one criterion consistently produced highest retention values.