BACKGROUND Potential differences in toxicity, potency and acquisition price among the liposomal amphotericin B formulations makes it unclear which agent is less costly when total resource consumption and treatment-associated costs are considered. METHODS A retrospective cost-minimisation analysis in 51 patients was performed to compare the cost of amphotericin B lipid complex (ABLC) and liposomal amphotericin B (L-AMB) from the hospital perspective. Costs ($US, 2001 values) were divided into level I (acquisition price only), level II (costs of all associated treatment, i.e. adverse events, failures, etc.) and level III (total fungal-related hospitalisation) costs. RESULTS No significant differences in patient demographics or length of therapy were apparent among those receiving ABLC or L-AMB. The clinical success rate in this population was similar between ABLC and L-AMB (53% vs 60%, p = 0.68), thus justifying the use of a cost-minimisation analysis. Among patients with baseline elevations in serum creatinine, 47% receiving ABLC and 10% receiving L-AMB experienced further increases in serum creatinine (p = 0.025). No differences in total treatment costs (level I, II, or III) were evident between patients receiving ABLC or L-AMB. When adjusted for duration of therapy, however, costs were significantly lower for ABLC than for L-AMB (level I: ABLC $US340 versus L-AMB $US435, p = 0.002; level II: ABLC $US361 versus L-AMB $US454, p = 0.027). The costs attributable to the prevention or treatment of adverse events were not different between the two treatments, and the economic outcome in this analysis was highly sensitive to the acquisition price and dosage of the lipid antifungal formulation. Two-way sensitivity analysis revealed that as long as the milligram price of L-AMB was greater than 135% of the milligram price of ABLC, ABLC remained the less costly formulation. CONCLUSIONS In this patient population, total hospitalisation costs were not different between lipid antifungal formulations. However, after controlling for duration of therapy, ABLC was less costly than L-AMB, when considering acquisition costs of the lipid antifungal agent and costs associated with concomitant antifungal therapy and the treatment of adverse events or lipid failures, indicating that the acquisition price of these agents should be predictive of their cost differences.