The forage ratio (FR) and Ivlev's electivity index (E) are commonly used to quantify food selection, but they depend on the relative abundances of food types in the environment, making them unsuitable for comparisons between food types with different abundances or for studying the relationship between selection and abundance. Modified versions of these indices are proposed, which are based on the rates of decrement (mortality) of food due to feeding and are independent of relative abundance. Selective feeding occurs when a feeder consumes co-occurring food sources at different rates. If $ m_A $ and $ m_B $ are the rates of mortality of food A and B due to feeding, $ N_A $ and $ N_B $ are the numbers of items of A and B in the environment, and t is time, then positive selection of A occurs if $ m_A = \frac{dN_A}{dt \cdot N_A} > m_B = \frac{dN_B}{dt \cdot N_B} $. Negative selection occurs if $ m_A < m_B $. There is no selection if $ m_A = m_B $. FR is defined as the ratio of the fraction r of a given food type in the feeder's ration to the fraction p of the same food in the environment: $ FR = \frac{r}{p} $. E is defined as the relative difference between r and p: $ E = \frac{r - p}{r + p} $. It is shown that $ FR = \frac{m_A}{m_T} $ and $ E = \frac{m_A - m_T}{m_A + m_T} $, where $ m_T $ is the total mortality rate of A and B. Both FR and E have a serious disadvantage: at constant mortality rates of A and B, their values change with the relative abundance.The forage ratio (FR) and Ivlev's electivity index (E) are commonly used to quantify food selection, but they depend on the relative abundances of food types in the environment, making them unsuitable for comparisons between food types with different abundances or for studying the relationship between selection and abundance. Modified versions of these indices are proposed, which are based on the rates of decrement (mortality) of food due to feeding and are independent of relative abundance. Selective feeding occurs when a feeder consumes co-occurring food sources at different rates. If $ m_A $ and $ m_B $ are the rates of mortality of food A and B due to feeding, $ N_A $ and $ N_B $ are the numbers of items of A and B in the environment, and t is time, then positive selection of A occurs if $ m_A = \frac{dN_A}{dt \cdot N_A} > m_B = \frac{dN_B}{dt \cdot N_B} $. Negative selection occurs if $ m_A < m_B $. There is no selection if $ m_A = m_B $. FR is defined as the ratio of the fraction r of a given food type in the feeder's ration to the fraction p of the same food in the environment: $ FR = \frac{r}{p} $. E is defined as the relative difference between r and p: $ E = \frac{r - p}{r + p} $. It is shown that $ FR = \frac{m_A}{m_T} $ and $ E = \frac{m_A - m_T}{m_A + m_T} $, where $ m_T $ is the total mortality rate of A and B. Both FR and E have a serious disadvantage: at constant mortality rates of A and B, their values change with the relative abundance.