OBJECTIVES Health Information Exchange (HIE) is emerging in healthcare markets across the US at an accelerating rate. However, one of the greatest barriers to building and sustaining HIE organizations is a clear understanding of the costs. We analyzed personnel costs for managing data connections in the Indiana health information exchange (IHIE). IHIE's data infrastructure is managed by project management (PM), system engineering (SE) and data mapping (DM) teams. Each team is involved in different tasks but to connect a hospital they coordinate overlapping tasks. To gage ongoing maintenance expenses we evaluated the cost differences among the teams. METHODS We interviewed team members to gather level of effort data for implementing and maintaining IHIE data interfaces. Personnel costs were calculated by multiplying the level of effort by staff salaries. Annual personnel costs were determined by the total interfaces implemented per year. We projected personnel costs for years 2004-2008 by using a constant level of effort and an increasing number of hospitals and annual salary. We used one-way ANOVA to detect differences among the average cost per team for managing interfaces. RESULTS The three teams (PM, SE, and DM) exhibited different average costs. The DM team produced the largest average cost and level of effort followed by SE and then PM. Using ANOVA and with critical value of 0.10, the cost differences were statistically significant (p=0.055). CONCLUSIONS Since the cost differences are significant, it suggests the differences are not due to chance but are largely due to level of effort variances. The level of effort differs because each team performs different tasks with varying degrees of complexity. Additional, difficult to quantify factors include size of hospitals, proficiency of hospital maintenance staff and adequate frequency of interactions; they appear in the average costs indirectly.