Health interventions for elderly people must understand the association between physical activity, sociodemographic factors, and non-communicable diseases. This study aimed to verify the association between physical activity in leisure time, sociodemographic factors, and NCDs in Brazilian older people. This is a descriptive study, with a cross-sectional design, carried out using secondary data from the 2019 National Health Survey—Brazil. It used data from 23,144 subjects aged over 60 years. Statistical analyses included descriptive and quantile regression with cutoff points 0.25, 0.50 (Median), 0.75, and 0.90 to verify the association between the variables. The statistical software R version 4.1.3 was used. Older people aged between 60 and 79 years were highlighted up to the 50th quantile. Females showed lower values in all quantiles, with emphasis from quantile 50 onwards. Subjects who self-declared as “white” showed significant differences up to quantile 50, not indicating significant values from this cutoff point. Residents of the rural area had lower values than residents of the urban area in all quantiles, with emphasis starting from the median. For cardiovascular diseases, it was found that subjects who did not report having this type of disease had better results for the amount of physical activity during leisure time, especially from the 75th quantile. It was concluded that there is a direct relationship between time spent in leisure-time performing physical activity, sociodemographic variables, and NCDs. It is necessary to review and validate cutoff points according to each category viewed, which can favor the adjustment of interventions according to each population. Actions of this nature can favor adherence by groups of older people to the weekly practice of physical activity.