This research was aimed to study the problems related to fish farming and the risks that they face.  Diyala Governorate was chosen as an applied model for measuring risk facing fish farmers. Research data were collected according to the random sample method, which included (30%) of all fish farming projects in Diyala Governorate. This sample has included 28% of the total fish farming projects in fish ponds and (31.5%) of the total Fish farming projects in cages through a questionnaire prepared for this purpose. The study focused on analyzing the behavior of fish farmers of the research sample towards risks by adopting a safety model in the form of the risk avoidance criterion (K_(s)(.The production functions for fish farming projects in fish ponds and floating cages were estimated, and these functions were estimated by the absence of dummy variables (educational level, years of experience) and with their presence, as well as estimating the risk functions of fish farming projects in fish ponds and floating cages. The results showed that the percent of fish farmers who prefer to take risks in fish ponds (25%) and in floating cages (33.3%), and that the percent of fish farmers who prefer moderate risk (natural or medium) were about (53%) in the ponds and the proportion (66.7%) in Cages, while the fish farmers who avoid risks, they constitute (22%) in ponds and (0%) in cages, from which we conclude that fish farmers in floating cages prefer a greater risk than fish farmers in fish ponds, also the results of the risk function for farming projects  in fish ponds and floating cages revealed that dummy variables (educational level, years of experience) have a negative relationship with production risks, which means that the improvement of the educational level and the increment of experience years contributes to a decrease in production risks.