Mortality due to poisonous snakebite is an emerging public health concern. Of late, snakebite has drawn global attention. Snakebite is a well-known cause of renal pathology especially in the haemotoxic group. Deaths due to renal failure are very common in envenomation due to poisonous snakebite. In this investigation our aim was to devise autopsy-proven diagnostic criteria of snakebite from the renal changes to confirm and classify the type of snakebite in forensic casework. That would be of practical help to the forensic pathologist with access to limited resources. An arbitrary scoring system from histology was adopted for the present semi-quantitative analysis. The scoring system (with scores from 0 to 5) developed by the present study was effective, valid and was able to distinguish between snakebite and control. The score obtained at histology in cases of snakebite was significantly different from the control group (Mann-Whitney U = 71.5; P = 0.011). Using stepwise linear discriminant function (DF) analysis, the score (Mukhopadhyay's score [MS]) was seen to be the only predictor of the type of snakebite. This model (DF = 1.05 x MS-1.78) could correctly classify 93.3% of the cases into neurotoxic and haemotoxic variety of offending snake. This is an effective, cheap and easy semi-quantitative method of distinguishing neurotoxic and haematoxic snakes from the renal changes at autopsy. It would be useful for practical application in centres where sophisticated biochemical or immunological analyses are not available due to limited resources.