Project Summary/Abstract Testing high volumes of clinical specimens for infectious diseases requires the use of efficient testing approaches. One approach used by laboratories is a procedure known as group testing (also known as pooled testing). In its most basic application, portions of specimens from different people are combined together into “groups” so that each corresponding individual is represented within one group. These groups are tested as if they were only single specimens. Members of negative groups are declared negative. Members of positive groups are retested separately in a second stage of testing to determine who is positive and who is negative. When group sizes are chosen in a statistically appropriate manner, the number of people represented by negative groups is much larger than those in positive groups. This leads to significant reductions in the overall number of tests required when compared to testing each specimen separately. These reductions subsequently result in significant increases for laboratory testing capacity by applying the resources saved to test more specimens. Current applications of group testing include: 1) testing blood donations for viruses, including hepatitis B and West Nile; 2) screening for bacteria that lead to chlamydia and gonorrhea; 3) checking for antiretroviral treatment failure among HIV-positive individuals; and 4) testing for viruses during a pandemic, including SARS-CoV-2. There are different algorithmic approaches to group testing. Members of positive testing groups can be successively split into smaller groups over two or more stages of testing. Alternatively, individual specimens can be allocated to multiple groups during the initial stage of testing in an effort to reduce the number of subsequent stages of testing. The first goal of this research to develop new group testing strategies that require few stages. This will enable laboratories to more easily implement group testing and to report test results quicker. The second goal is to develop new statistical learning methods for data arising through group testing. These methods will result in better predictions for the probability of positivity and can be used to develop more efficient approaches to implement group testing. The third goal is to create tools for laboratories so that they can apply this research. These tools will include a web-based application that allows laboratories to choose the most efficient group testing strategy for their particular situation.