This study aims to research the causality relationship between industrial agglomeration in labor, fisheries production, and fisheries households in Tanggamus district. The agglomeration index measured using the Balassa Hoover Index analysis tool and Geographic Information System (GIS) using Geoda. The data used are secondary data obtained directly from the Central Statistics Agency (BPS), the Department of Fisheries and Maritime Affairs of Tanggamus Regency. The number of cross-section data is 20 districts in the period 2012-2016. analyzed using the Granger Causality Analysis method. The Hoover Ballassa index results show a strong degree of agglomeration in only 6 districts. The results of the analysis of the relationship between states of origin state that there is a unidirectional causality between the agglomeration and labor variables, which statistically significantly influence the agglomeration and do not apply otherwise. Unidirectional causality occurs between agglomeration variables and fisheries production only fisheries production which statistically significantly influences agglomeration and does not apply vice versa. Unidirectional causality occurs between agglomeration variables and fisheries households ie only fishery households that statistically significantly influence agglomeration and do not apply otherwise.