Nitric oxide (NO) releasing agents have, in experimental settings, been shown to decrease intravascular nitrogen bubble formation and to increase the survival rate during decompression sickness (DCS) from diving. The effect has been ascribed to a possible removal of preexisting micronuclei or an increased nitrogen washout on decompression through augmented blood flow rate. The present experiments were conducted to investigate whether a short- or long-acting NO donor [glycerol trinitrate (GTN) or isosorbide-5-mononitrate (ISMN), respectively] would offer the same protection against spinal cord DCS evaluated by means of spinal evoked potentials (SEPs). Anesthetized rats were decompressed from a 1-h hyperbaric air dive at 506.6 kPa (40 m of seawater) for 3 min and 17 s, and spinal cord conduction was studied by measurements of SEPs. Histological samples of the spinal cord were analyzed for lesions of DCS. In total, 58 rats were divided into 6 different treatment groups. The first three received either saline ( group 1), 300 mg/kg iv ISMN ( group 2), or 10 mg/kg ip GTN ( group 3) before compression. The last three received either 300 mg/kg iv ISMN ( group 4), 1 mg/kg iv GTN ( group 5), or 75 μg/kg iv GTN ( group 6) during the dive, before decompression. In all groups, decompression caused considerable intravascular bubble formation. The ISMN groups showed no difference compared with the control group, whereas the GTN groups showed a tendency toward faster SEP disappearance and shorter survival times. In conclusion, neither a short- nor long-acting NO donor had any protective effect against fatal DCS by intravenous bubble formation. This effect is most likely due to a fast ascent rate overriding the protective effects of NO, rather than the total inert tissue gas load.