In recent years an increasing number of publishers have adopted tiered pricing of journals. The design and implications of tiered-pricing models, however, are poorly understood. Tiered pricing can be modeled using several variables. A survey of current tiered-pricing models documents the range of key variables used. A sensitivity analysis identified two main variables: the distribution of the subscriber base across tiers and the price differential between the highest and lowest tiers. Scenarios illustrating the effects of these two variables are presented and analyzed. The results suggest that institutions falling into the highest tiers can experience substantial price increases, particularly if the subscriber distribution is weighted toward the lower tiers of the model. The implications for collection management at institutions of all sizes are considered. (Contains 5 table and 12 notes.)