This thesis advances the way economic values for livestock trait improvement are obtained by taking into account risk. This thesis also investigates and contrasts two ways of obtaining economic values, namely economic values without consideration of risk (traditional economic values) and economic values which incorporate variance of profit and risk attitudes of livestock producers (risk-rated economic values). Traditional economic values were derived using a profit model, whereas risk-rated economic values were derived using a risk-rated profit model (a certainty equivalent profit model). The major distinctions in the derived formulae for economic values were in the fact that the risk-rated economic values included variance of profit and the Arrow-Pratt coefficient of absolute risk aversion. Risk-rated economic values differed, in absolute terms, from the traditional economic values, in both cases with and without optimisation of the management choice variable. For different values of the Arrow-Pratt coefficient of absolute risk aversion, the differences between traditional and risk-rated economic values increased as the risk aversion and variance of output price increased. The differences between economic values, in absolute terms, were greater when optimisation of the management choice variable was considered. The above differences would have an effect on the cost-benefit analysis of breeding programs and might affect directions for selection or emphasis on traits. When the relative economic values were considered (the ratios of economic values), the differences between the traditional economic values (derived using the profit model) and risk-rated economic values (derived using the risk-rated profit model) were greater than 50% without and with optimisation of management, in some cases that were considered. A practical example of deriving and comparing traditional and risk-rated economic values was considered for selection in a population based on use in the feedlot phase of beef production. One output (average daily gain) and one input (daily dry matter intake) traits were considered. Traditional economic values and risk-rated economic values differed in both absolute and relative terms. The use of traditional economic values when the objective was certainty equivalent profit resulted in underestimation of response. The overall conclusion, based on the practical example considered, is that, in some cases, it is important to consider risk when deriving economic values. Consideration of risk can effect cost-effectiveness of selection.