The coronavirus pandemic changed paratransit service dramatically, with most operators eliminating shared rides to halt disease transmission. This paper applies an estimate of disease contact exposure to actual data from New York City’s Access-A-Ride (AAR) paratransit system. Scenario analyses performed using insertion heuristic trip construction showed that eliminating rideshare on the AAR system increased operating miles by 70%. We also showed contact exposure can be significantly limited (by;50% to 60%) by reducing vehicle capacity from 4 to 2 passengers. Manipulating the maximum ride time factor also showed the potential to reduce contact exposure, with a 50% increase of the ratio from 2 to 3 leading to a 40% increase in contact exposure. Contact exposure was relatively insensitive to changes in the maximum wait time policy.