Replacement decisions are influenced by various factors. Empirical values for some of these factors were determined. It was found that replacement costs can be forecasted with an irregular curve, while trade-in sales can be determined with a regression function based on new cost price, age and tractor power.