The research identified the need to improve mathematical models of the machinery milking duration, which will contribute to increasing the accuracy of assessing the productivity of milking plants at the stages of their development and modernization. Also, the development of these models will allow to further create the methodology for designing information systems to support and ensure the technological process of milk production in the conditions of various types of dairy farms. The study examines the basic algorithms of modern milking machines. Based on the experimental data obtained, it was found that the duration of animal preparation for milking (if milking robots are not used) has a chi-square distribution. The duration of milking an animal when using the milking machine without a milking process control function, and the duration of a random milking time interval when using the milking machine with a milking control function, have a gamma distribution. Mathematical models of the machinery milking duration when using the milking robot, the milking machine without a milking process control function, and the milking machine with a milking process control function have been improved, which take into account the minimum operating time of the milking machine and the time period during which the animal is identified.