The Flow Monitoring component of the DTM (Displacement Tracking Matrix)1 tool consists of a registry of migrants and the application of surveys directed at representatives of travel groups, facilitating the generation of quantitative estimates on the flow of people in a defined location and, from this, the collection of qualitative data. The data for the flow monitoring registry is obtained from official data provided by the Government of Panama, corresponding to the voluntary departures of migrants from the MRS of Los Planes, in Gualaca, through the National Migration Service (SNM). The instrument used for the flow monitoring surveys is in Spanish, English, Haitian Creole, French and Portuguese in order to increase the number of people who can be surveyed and collecting data on gender, age, nationality, characteristics of the trip, migration route, country of habitual residence, in addition to counting the people who make up the travel group of the people surveyed. For the definition of monitoring sites (Map 1), the baseline for the evaluation of irregular migration flows and mobility tracking in border localities in southern Costa Rica under the context of COVID-19 was taken as a reference, based on consultations with key informants2 and field work. For the study, records of irregular departures from Panama between April 1 and April 30, 2022, were obtained, corresponding to 5,136 irregular departures to Costa Rica. In parallel, a stratified random sampling by monitoring points was carried out during April. A total of 133 surveys were conducted with migrants, 54 per cent (72) were valid surveys and 46 per cent were invalid (refusals by respondents). The data analysis was based on the assignment of the sample weight, which is useful for extrapolating the characteristics of the people interviewed to the rest of the people recorded in the exit registers. With this statistical analysis it is possible to know, by means of estimates, the characteristics of the total number of people counted. Of the 133 surveys carried out, 61 responses were obtained in which informed consent was not accepted; therefore, when weighting the sample weight, they were excluded from the calculation.