Pick excavatability of controlled low-strength materials was experimentally studied 10 to circumvent the fact that current classification systems or empirical excavatability criteria lack relevance (they do not refer to digging facilities and depend on the mix design and the granular skeleton of the material). Manual excavation is separated into two steps: the impact of the pick and decompaction. This was analysed by a mechanical approach. Energy is mainly consumed during the first step. A laboratory punching test was therefore proposed to evaluate excavatability (a confined cylindrical specimen was loaded by a cylindrical punch). Results were compared between the mechanised and the manual pick excavation tests on field trenches backfilled with air-entrained mortars or very lean concretes. These comparisons confirmed that a limited compressive strength was an indicator of excavatability. A model was proposed to evaluate the excavated volume by punching penetration for a given level of energy and a maximum size of aggregates in the material