The results of a four-year longitudinal study of the development of processing speed throughout primary school years are presented. The task of the study is to analyze the average path of development of the processing speed at primary school age, to assess individual-specific deviations from the average path for each participant in the study, as well as to identify gender differences in average values, the path of development of the speed of information processing and groups of children with different severity reaction times - slow and fast. The sample consisted of 224 schoolchildren (46% of girls) of grades 1-4 of the educational institution participating in the Russian longitudinal study of the academic success of schoolchildren. The average age of schoolchildren during the first measurement (first grade) was 7.85 years (standard deviation = 0.34) and 10.77 years (standard deviation = 0.36) during the fourth measurement (fourth grade). Participants completed the computerized task, "The Response Time of Choice," at the end of each school year throughout the entire period of primary education. To study the developmental trajectory of the processing speed, we used the growth analysis method with mixed effects, which allows us to estimate the average trajectory for the entire sample and individually-specific deviations from the average trajectory for each person. It is shown that across primary school years, the reaction time decreases, and the speed of processing, respectively, increases. At the same time, the trajectory of the reaction time changes nonlinearly: the most intense decline is observed from the first to second year of education in elementary school, then - until the fourth grade, the rate of decline slows down. Boys are ahead of girls in average processing speeds for each year of primary school education. However, the pace and nature of changes in the speed of processing are not statistically significantly different for boys and girls during across primary school years. To analyze gender differences in the processing speed in groups of schoolchildren with different reaction times, a quantile regression analysis was used to evaluate differences at opposite ends of the distribution of the variable. It is shown that girls and boys do not differ in the speed of information processing in groups of fast and slow primary school children.