Freshwater mussels are particularly susceptible to injury from exposure to hazardous substances due to their sessile nature and filter feeding biology. There have been various Natural Resource Damage Assessment and Restoration (NRDAR) cases in the United States involving injury to freshwater mussels due to releases of hazardous substances into rivers and streams. Restoration of mussels in these cases typically involves propagation of mussels at a hatchery facility and their subsequent stocking or release at restoration sites. However, determination of the services lost due to injury to mussel populations and the appropriate level of restoration (and associated costs) to recover those losses has varied among NRDAR cases. Standardized methods would facilitate injury determination and restoration planning for future cases involving injury to mussels. The purpose of this research was to use two of the earliest and largest NRDAR cases (Certus, Inc. and Lone Mountain Processing, Inc. (LMPI)) involving injury to mussels to: 1) determine whether restoration for these cases was sufficient and 2) analyze restoration efforts for application in future NRDAR cases (i.e., lessons learned and development of standardized methods). This study represents the first evaluation of mussel restoration efforts in a NRDAR context. In general, 4.8% to 6.1% of juvenile mussels that excysted from host fishes in the hatchery survived to be eventually released at restoration sites. Further, based on expected survival and recruitment rates of released mussels, monitoring of restoration sites found 43% to 15% of the expected number of mussels. Understanding reasons for this discrepancy between expected and estimated survival is critical for determining the level of restoration success. If released mussels are either establishing and/or recruiting outside of monitoring area but otherwise alive and breeding, then they should count towards successful restoration. In contrast, if released mussels have either high mortality over time or are dying shortly after release, then expected gains from these mussels should not count towards successful restoration. I developed a mussel-specific Resource Equivalency Analysis (REA) for use in future NRDAR cases that compares the loss of services, using Discounted Mussel Years (DMYs) as units, to the expected gain in services from restoration. Applying this analysis to the Certus and LMPI NRDAR cases suggests that mussel restoration was successful (i.e., expected DMYs gained are greater than those lost), even when it was assumed that 75% of released mussels were dying after being released at restoration sites. Finally, a cost analysis of two mussel propagation facilities found that the yearly cost per mussel released at a restoration site ranged from $4.36 to $96.48. The suite of species propagated each year varied. As some species are more difficult to propagate than others, the cost per mussel varied widely. These data will facilitate the determination of restoration costs in future cases. Together, this information provides a starting point for consistently estimating restoration effort and costs for future NRDAR cases involving freshwater mussels. / Doctor of Philosophy / Freshwater mussels provide numerous ecosystem services. Most importantly, they purify large volumes of water, and provide habitat and food for other animals. However, they are highly vulnerable to chemical spills because they cannot move long distances quickly and they are directly exposed to toxic substances if they filter water. There have been many cases in past decades where vulnerable mussel populations were exposed to chemical spills. When these populations are injured, the services they provide are lost until the population can be restored to pre-spill conditions. Restoration of mussel populations usually involves raising juvenile mussels in hatchery facilities and then releasing them in areas where populations were injured. Determining the appropriate level of restoration needed to restore populations has varied widely among cases. A standardized approach would facilitate determination of restoration and restoration costs. I used data from two cases (Certus, Inc. and Lone Mountain Processing, Inc.) where mussel populations were injured due to a chemical spill to: 1) determine whether restoration for these two cases was successful and 2) develop tools and draw insights for use in future cases where mussels are injured. This study represents the first evaluation of restoration success of freshwater mussels in a NRDAR context. On average, 4.8% to 6.1% of juvenile mussels produced at two Virginia hatchery facilities survived to be released at restoration sites. Further, of the mussels released, only 43% to 15% of the expected mussels were found in later years. These "missing" mussels are either leaving and/or breeding outside of their release areas, or they are dying and failing to provide important ecosystem services. Further study is needed to determine the degree to which each of these is the case. I also developed a mussel-specific method of determining how much restoration is needed to provide the amount of ecosystem services as pre-spill conditions (called Resource Equivalency Analysis or REA). Application of REA to these two test cases (Certus, Inc. and Lone Mountain Processing, Inc.), I showed that restoration for these cases was successful, even if as much as 75% of released mussels are dying after being released at restoration sites. Finally, I found that the cost of successfully releasing a mussel ranged from $4.36 to $96.48 per mussel. This information is useful for estimating the cost of restoration plans in future chemical spills that injure freshwater mussels.