In vitro assays are normally conducted in plastic multi-well plates open to exchange with the ambient air. The concentration of test substances freely available to cells is often not known, can change over time, and is difficult to measure in the small volumes in microplates. However, even a well-characterized toxicological response is of limited value if it cannot be linked to a welldefined exposure level. The aim of this study was to develop and apply an approach for determining time resolved freely dissolved concentrations of semi-volatile and hydrophobic organic chemicals (SVHOCs) in in vitro assays: (1) free fractions were measured by a new medium dilution method and (2) time-resolved loss curves were obtained by measurements of total concentrations in 96-well plates during incubations at 37°C. Headspace solid-phase microextraction was used as an analytical technique for 24 model chemicals spanning 6 chemical groups and 4-5 orders of magnitude in K ow and K aw. Free fractions were above 30% for chemicals with log K ow < 3.5 and then decreased with increasing log K ow. Medium concentrations declined significantly (>50%) within 24 hours of incubation for all 20 chemicals having log K ow > 4 or log K aw >-3.5 in serum free medium. Losses of chemicals were lower for medium containing 10% fetal bovine serum, most significantly for chemicals with log K ow > 4. High crossover to neighboring wells was observed also below log K ow of 4 and log K aw of-3.5. Sealing the well plates had limited effect on the losses, but clearly reduced crossover. The high losses and crossover of most tested chemicals question the suitability of multi-well plates for in vitro testing of SVHOCs, and call for (1) test systems that minimize losses, (2) methods to control in vitro exposure, (3) analytical confirmation of exposure and (4) exposure control and confirmation being included in good in vitro reporting standards.