Tungiasis is a parasitic skin disease prevalent in impoverished populations in the tropics and associated with considerable morbidity. Treatment options are limited and prophylaxis has never been attempted. We assessed the effectiveness of a plant-based repellent to prevent infestation with Tunga penetrans and sand flea disease in an urban squatter settlement in Fortaleza, Northeast Brazil. Two cohorts were formed. One cohort started with the intervention while another served as control. The plant-based repellent Zanzarin was applied twice daily on the skin of the feet. After 4 weeks and a wash-out-phase of another 2 weeks, cross-over was performed. Study members were examined twice per week for the presence of newly embedded sand fleas. The number of viable and dead sand fleas and the total number of sand flea lesions were determined. Sentinel rats were used to assess local transmission dynamics. Outcome measures were the number of newly embedded sand fleas per individual per day, number of sand flea lesions per individual according to the stage, as well as type and degree of clinical pathology. Zanzarin applied twice daily on the feet reduced the number of newly embedded fleas by 92% (interquartile range 81-100%), the total number of lesions by 87% (81-98%) and reversed tungiasis-associated clinical pathology almost completely. In conclusion, in a setting in which tungiasis is an important health threat, the daily application of a plant-based repellent prevented the infestation with T. penetrans and sand flea disease.