Food safety regulations are evolving to more performance-based regimes in which firrms have greater flexibility and responsibility for adopting effective controls. Within this context, this paper compares performance and process standards modeling the variability of industry-level compliance and therefore the resultant level of food safety. Monte Carlo simulations are conducted manipulating five factors: the variances of input use of efficient and inefficient firms, the proportion of inefficient firms in the industry, the mean of the error term for inefficient firms, and the policymakers' level of risk aversion. Results suggest that process standards may be preferred over performance standards when inefficient firms prevail in the industry, when input use is highly variable and when the regulator pays less attention to the variability of industry-level compliance and more to the level of the standard.