The current method for Paralytic Shellfish Poisoning (PSP) testing in shellfish is based on the mouse bioassay (MBA), which involves injecting shellfish extract into a conscious mouse, and then converting its time to death into PSP toxicity using Sommer's table. To improve animal welfare, the present study investigated the use of anaesthesia. A saxitoxin (STX) calibration study was conducted where known amounts of STX were injected into both unanaesthetised and anaesthetised mice. Death time was approximately doubled when mice were anaesthetised. Both unanaesthetised and anaesthetised animals showed a linear relationship between the inverse death time and log(STX). Based on these data, new calibration curves were developed. This study revealed that the current method employing Sommer's table underestimates toxicity by up to 50% for higher toxin levels. Subsequently, shellfish samples were tested on both unanaesthetised and anaesthetised mice. Using the new calibration curves, the numbers of samples exceeding the field closure limit were similar for unanaesthetised and anaesthetised mice, and were nearly two-fold higher than those obtained with the current method. The studies showed that the bioassay gives variable results for both unanaesthetised and anaesthetised animals. Anaesthesia forms a viable and more ethical alternative to the current bioassay, at least in the short term. A practical summary on how to conduct this method is given.