In a study, Kawai et al., test the use of a new gelatin glue as a dural sealant in in vitro and in vivo canine models of transdural CSF leakage. The in vitro model was sutured semicircles of canine dura mater and artificial dural substitute. The sutures were sealed with gelatin glue (n = 20), fibrin glue (n = 20), or a polyethylene glycol (PEG)-based hydrogel sealant (n = 20). Each sample was set in a device to measure water pressure, and pressure was increased until leakage occurred. Bonding strength was subjectively evaluated. The in vivo model was dogs who underwent dural excision and received either no sealant (control group; n = 5) or gelatin glue sealant (n = 5) before dural closure. Twenty-eight days post-surgery, the maximum intracranial pressure was measured at the cisterna magna using Valsalva maneuver and tissue adhesion was evaluated. The water pressure at which leakage occurred in the in vitro model was higher with gelatin glue (76·5 ± 39·8 mmHg) than with fibrin glue (38·3 ± 27·4 mmHg, P < 0·001) or the PEG-based hydrogel sealant (46·3 ± 20·9 mmHg, P = 0·007). Bonding strength was higher for the gelatin glue than fibrin glue (P < 0·001) or PEG-based hydrogel sealant (P = 0·001). The maximum intracranial pressure in the in vivo model was higher for the gelatin glue group (59·0 ± 2·2 mmHg) than the control group (13·8 ± 4·0 mmHg, P < 0·001). Tissue adhesion was lower for the gelatin glue group than the control group (P = 0·005). The new gelatin glue provides an effective watertight closure when used as an adjunct to sutured dural repair 1).