Skeletal soils are not suitable for agriculture, and often are allocated to marginal uses such as cherry orchards for timber production. These require some agricultural practices (irrigation, soil tillage or weed control) which can contribute to the development of a hardpan. Compacted layers can adversely affect timber production, so subsoiling works are required. This study examined the effect of six years of tillage on hardpan formation in a skeletal soil by means of mechanical impedance measurements with a dynamic penetrometer cone (dynamic cone test), a method that is quick and easy to use, but can suffer from interference by stones. Mechanical impedances along the soil profile were measured in four plots differing in tillage (with or without) and drip irrigation (with or without) treatments. Exploratory data analysis together with statistical inference techniques related to linear general models were applied. The presence of a transitional layer on top of the hardpan is suggested in the non-tilled plot and soil depth that can be explored easily by roots has increased by 20 cm.