Minimal-exchange, intensive biofloc aquaculture systems offer a viable means of culturing marine animals at inland locations due to very low rates of water use. Fresh, never-frozen shrimp can be provided to metropolitan markets; however, the cost of artificial salt can be substantial. The purpose of this project was to examine commercial-scale biofloc shrimp production at three different salinities. Nine raceways were randomly assigned to three salinity treatments: 10, 20, and 30‰ (LS, MS, and HS), each treatment contained three raceways operated at 50 m³. The raceways were operated as heterotrophic biofloc systems, with daily additions of sucrose to raise the C:N ratio. Temperature, dissolved oxygen, pH, and salinity were all maintained at consistent levels. Spikes of ammonia and nitrite occurred in all tanks but nitrate remained low, with a peak value of 8.7 mg NO3-N L− 1. There were no significant differences in any shrimp production metric. Mean shrimp growth rate was 1.8, 2.0, and 2.0 g week− 1 in the LS, MS, and HS treatments respectively. Mean feed conversion rate was 1.6, 1.2, and 1.2 in the LS, MS, and HS treatments respectively, and mean final weight ranged from 17.8 to 19.3 g. The only time water was removed from the systems was when settling chambers were emptied, resulting in a total mean water replacement of 5.2% or less per raceway. The mean volume of full strength seawater used to produce shrimp was 104, 159, and 235 L kg− 1 of shrimp in the LS, MS, and HS treatments respectively. Although there were no significant differences in shrimp production metrics between treatments, these values were noticeably lower in the LS treatment due to human error. Operating at the low salinity of 10‰ reduces salt use by about 50% over the MS treatment which implies substantial cost savings for production facilities. This study helps to illustrate the range of salinity options for shrimp production in commercial-scale biofloc systems.