The Ozone Water-Land Environmental Transition Study, 2018 (OWLETS-2) measured total non-methane hydrocarbons (TNMHC) and EPA PAMS Volatile Organic Compounds (VOCs) on an island site in the northern Chesapeake Bay 2.1 and 3.4 times greater in concentration, respectively, than simultaneous measurements at a land site just 13 km away across the land-water interface. Many PAMS VOCs had larger concentrations at the island site despite lower NEI emissions over the water, but most of the difference comprised species generally consistent with gasoline vapor or exhaust. Sharp chemical differences were observed between the island and mainland and the immediate air ~300 m above the water surface observed by airplane. Ozone formation potential over land was driven by propene and isoprene but toluene and hexane were dominant over the water with little isoprene observed. VOC concentrations over the water were noted to increase diurnally with an inverse pattern to land resulting in increasing NOx sensitivity over the water. Total reactive nitrogen was lower over the water than the nearby land site, but reservoir compounds (NOz) were greater. Ozone production rates were generally slow (~5 ppb hr-1) both at the surface and aloft over the water, even during periods of high ozone (>70 ppbv) at the water surface. However, specific events showed rapid ozone production >40 ppb hr-1 at the water's surface during situations with high VOCs and sufficient NOx. VOC and photochemistry patterns at the island site were driven by marine sources south of the island, implicating marine traffic, and indicate ozone abatement strategies over land may not be similarly applicable to ozone over the water.Implications: Measured chemical properties and patterns driven primarily by marine traffic sources over water during ozone conducive conditions were starkly different to immediately adjacent land sites, implying ozone abatement strategies over land may not be similarly applicable to ozone over the water.