Chemical immobilization is an effective technique to suppress the release of arsenic from naturally arsenic-containing excavated rock/sediment. For designing the chemical immobilization technique, it is important to understand that the immobilization of arsenic depends on the sizes of ionic arsenic and arsenic retained on the colloids and suspended particles that are released from the excavated rock/sediment. Tests on the size fractionation of the arsenic released and the subsequent immobilization were conducted. The total amount of the size fraction of arsenic released from six excavated rock/sediment ranged from 0.16 to 0.75 mg kg^-1. The distributions of size fraction of arsenic released were categorized into three types: the dominant fraction was suspended particle fraction (SP-F) and ionic fraction (I-F), and a compatible amount of SP-F and I-F was included. Steel slag, calcium oxide, and ferrihydrite, which can effectively and stably immobilize ionic arsenic with different mechanisms, decreased the total amounts of the size fraction of arsenic released at 28%-84%, 59%-83%, and 57%-84%, respectively. Ferrihydrite and calcium oxide greatly reduced the I-F and the small and large colloid fractions. The steel slag was effective in reducing the SP-F at >86 %. In most arsenic fractions, the immobilized arsenic was not re-released at <7 %. This study provides the first experimental evidence of the variation in the released arsenic size depending on the excavated rock/sediment. In addition, the size fraction of the arsenic that could be immobilized depended on the immobilizing material. Thus, it is suggested that the combined application of immobilization materials would present a useful approach for immobilizing various released arsenic phases and preventing immobilized arsenic from re-release.