For the first time, a Live/Dead® (L/D) Bacterial Viability Kit (BacLight™) protocol was adapted to marine sediments and applied to deep-sea sediment samples to assess the viability (based on membrane integrity) of benthic bacterial communities. Following a transect of nine stations in the Fram Strait (Arctic Ocean), we observed a decrease of both bacterial viability and abundance with increasing water (1250-5600 m) and sediment depth (0-5 cm). Percentage of viable (and thus potentially active) cells ranged between 20-60% within the first and 10-40% within the fifth centimetre of sediment throughout the transect, esterase activity estimations (FDA) similarly varied from highest (13.3±5.4 nmol cm-3 h-1) to lowest values below detection limit down the sediment column. Allowing for different bottom depths and vertical sediment sections, bacterial viability was significantly correlated with FDA estimations (p<0.001), indicating that viability assessed by BacLight staining is a good indicator for bacterial activity in deep-sea sediments. Comparisons between total L/D and DAPI counts not only indicated a complete bacterial cell coverage, but a better ability of BacLight staining to detect cells under low activity conditions. Time course experiments confirmed the need of a rapid method for viability measurements of deep-sea sediment bacteria, since changes in pressure and temperature conditions caused a decrease in bacterial viability of up to 50% within the first 48 h after sample retrieval. The Bacterial Viability Kit proved to be easy to handle and to provide rapid and reliable information. It's application to deep-sea samples in absence of pressure-retaining gears is very promising, as short staining exposure time is assumed to lessen profound adverse effects on bacterial metabolism due to decompression. © 2004 Elsevier B.V. All rights reserved.