Tumor necrosis factor alpha (TNF-α) is known to be crucial in many biological activities of organisms. In this study, physicochemical properties and modeling of TNF-α protein of fish was analyzed using in silico approach. TNF-α proteins selected from fish species, including grass carp (Ctenopharyngodon idella), zebra fish (Danio rerio), Nile tilapia (Oreochromis niloticus), goldfish (Carassius auratus), and rainbow trout (Oncorhynchus mykiss) were used in this study. Physicochemical characteristics with molecular weight, theoretical isoelectric point, extinction coefficient, aliphatic index, instability index, total number of negatively charged residues and positively charged residues, and grand average of hydropathicity were computed. All proteins were classified as transmembrane proteins. The “transmembrane region” and “TNF” domain were identified from protein sequences. The function prediction of proteins was also performed. Alpha helices and random coils were dominating in the secondary structure of the proteins. Three-dimensional structures were predicted and verified as good structures for the investigation of TNF- of fish by online server validation.