Three non-steroidal anti-inflammatory drugs (NSAIDs)-ketoprofen, naproxen and piroxicam-in both deionized (DI) water and surface lake water (SW) (Tallahassee, FL), were exposed to varying ozone treatment regimes or H 2 O 2 /O 3 advanced oxidation on the laboratory bench. Recently used biofilm-supporting granular activated carbon (BAC) was sampled from a municipal drinking water treatment facility (Tampa, FL, USA), and employed to determine the bio-availability of chemical intermediates formed in ozonated water. Advanced chemical analysis was used to identify oxidation by-products formed and combined with a bioanalytical tool to assess non-specific toxicity (Microtox assay). All 3 target pharmaceuticals were efficiently removed by different processes, with a lower NSAIDs removal yield observed in lake water compared to DI water experiments. The removal yields of ketoprofen, naproxen, and piroxicam improved with increasing ozone dose, H 2 O 2 /O 3 ratio and empty bed contact time (EBCT) with BAC. Ozonation with BAC filtration had a positive impact by reducing the initial ozone dose required to achieve > 90% removal of all 3 pharmaceuticals (when an initial ozone dose < 1 mg L -1 was combined with EBCT < 15 min). The toxicity evolution of the treated samples was monitored by Microtox bioassay. Ozone doses higher than 2 mg L -1 for 2 min contact time were optimal to reach the lower water samples toxicity with NSAIDs removal yields ranging from 95.5 to 99.0% in DI water and from 77 to 90% in SW. Also, higher ozone doses were not shown to remove the residual toxicity. In contrast, the BAC filtration hardly decreases the sample toxicity when an EBCT of 15 min was