Social contact patterns are key factors to model the spread of respiratory infections as well as to assess the impact of behavioral interventions such as isolation and social distancing. During the pandemic of SARS-CoV-2, different types of interventions aimed to affect individuals’ mixing patterns have been critical for the epidemic containment. Data on social contact patterns however are limited in low- and middle-income countries, where extremely diverse socio-demographic characteristics are present, even within a single country. Hence, the aim of this work, based on a survey of 1,407 individuals from rural, urban and slum settings in Kenya, is to provide an in depth analysis of social contact patterns in a low-income setting prior to COVID-19, and assess their determinants in social contexts that are highly diverse in terms of structural sociodemographic factors (e.g., setting, age, and household size) and behavioral factors (e.g., time activity patterns).