Agriculture operates in an ever changing environment which makes this sector vulnerable to a number of risks and uncertainties. Among the various risks farmers face, production risks (particularly catastrophic risks) presents the most dominant sources of risks and uncertainties in agriculture. Farmers use available risk management tools to mitigate/minimize the potential adverse impacts of such risks and uncertainties at farm level. This study highlight important factors (internal, external and behavioral factors) affecting risk management decisions at farm level and provide some methodological approaches for quantification of these variables and analyzing their effect on farmers’ decisions of adopting risk management tools. The risk management decisions depend on various factors which can be broadly categorized into internal and external factors. Internal factors include farm (farm size, ownership of land etc.) and farm household characteristics (gender, age, education, income, family size etc.). External factors consist of availability and access to information and credit sources and input/output markets. Besides the internal and external factors, there are some behavioral attributes that also effect farmers’ risk management decisions. Farmers’ risk perceptions and their attitude towards risks are significant factors affecting farmers’ decisions of adopting various risk management tools. Another important aspect of the decision process is the simultaneous adoption of multiple tools at the same time i.e. adoption of one risk management tool may make it more likely to adopt other available risk management tool(s). Various methodological approaches are used to quantify these variables and fetch some meaningful results from the field data. The internal and external factors are relatively easy to measure however eliciting farmers’ perceptions and their risk attitude is tricky and require sound methodological approaches. Perceptions can be recorded using a likert scale and can be processed using a risk matrix while Equally Likely Certainty Equivalent (ELCE) or Toss Method can be used to elicit risk attitude from an economic agent. Similarly the simultaneous adoption effect can be best captured using a multivariate probit/logit model.