Belgium’s Law of War Manual (1983) defines armed forces as comprising

A report submitted to the Belgian Senate in 1991 noted that two elements were essential in the definition of armed forces: first, they must be integrated into a military organization (that is, a hierarchical structure) subject to an internal disciplinary system; second, this organization must operate under a command structure responsible to a party for the conduct of its subordinates. If these two conditions were fulfilled, the concept of armed forces could be extended to groups of combatants who were left behind in an occupied territory to perform acts of sabotage, to gather intelligence or to take part in guerrilla warfare. The report recalled that this was the position of the Belgian government in exile during the Second World War. From its base in London, the government adopted legislation authorizing the executive power to nominate agents in charge of action or intelligence missions in a foreign country, occupied area or zone evacuated by the enemy. These agents had the status of combatants and were allowed to carry arms. The government in exile, however, was very reticent about resistance cells or individuals over whom it had no direct control. 

Upon ratification of the 1977 Additional Protocol I, Belgium notified the High Contracting Parties of the duties assigned to the Belgian Gendarmerie (constabulary) in time of armed conflict. Belgium considered that this notification fully satisfied any and all requirements of Article 43 pertaining to the Gendarmerie. It informed the High Contracting Parties that the Gendarmerie was formed to maintain law and order and was, according to national legislation, a police force which was part of the armed forces within the meaning of Article 43 of the 1977 Additional Protocol I. Consequently, members of the Gendarmerie had the status of combatant in time of international armed conflict.