With animal models, death as an intentional endpoint is ethically unacceptable. However, in the study of septic shock, death is still considered the only relevant endpoint. We defined eight humane endpoints into 4 stages of severity (from healthy to moribund) and used to design a clinically relevant scoring tool, termed 'the mouse clinical assessment score for sepsis' (M-CASS). The M-CASS was used to enable a consistent approach to the assessment of disease severity. This allowed an ethical and objective assessment of disease after which euthanasia was performed, instead of worsening suffering. The M-CASS displayed a high internal consistency (Cronbach's α = 0.97) with a high level of agreement, and an intraclass correlation coefficient (ICC) equal to 0.91. The plasma levels of cytokines and markers of oxidative stress were all associated with the M-CASS score (Kruskal Wallis test, p<0.05). The M-CASS allows tracking of disease progression and animal welfare requirements.