This article presents the results of a three-year experiment (2018–2020) conducted at the Agricultural Experiment Station in Bałcyny (north-eastern Poland) with the aim of determining Jerusalem artichoke (JA) yields and the energy balance of biomass production in (i) a perennial cropping system (only aerial biomass was harvested each year) and (ii) an annual cropping system (both aerial biomass and tubers were harvested each year). When JA was grown as a perennial crop, the demand for energy reached 25.2 GJ ha−1 in the year of plantation establishment and 12.3–13.4 GJ ha−1 in the second and third year of production. The energy inputs associated with the annual cropping system were determined in the range of 31.4–37.1 GJ ha−1. Biomass yields were twice as high in the annual than in the perennial cropping system (20.98 vs. 10.30 Mg DM ha−1). Tuber yield accounted for 46% of the total yield. The energy output of JA biomass was 1.8 times higher in the annual than in the perennial cropping system (275.4 vs. 157.3 GJ ha−1). The average energy gain in JA cultivation ranged from 140 (perennial crop) to 241 GJ ha−1 (annual crop). The energy efficiency ratio of JA biomass production reached 7.7–13.3 in the perennial cropping system, and it was 20% lower in the annual cropping system. These results imply that when JA was grown as an annual crop, an increase in energy inputs associated with plantation establishment (tillage and planting) and the harvest and transport of tubers was not fully compensated by the energy output of tubers.