Although producers’ prime objective may be to increase net returns, many are also interested in conserving and enhancing the quality the soil, water and air resources through adopting more environmentally friendly production practices. This study compared non-renewable energy inputs, energy output, and energy use efficiency of nine dryland cropping systems comprised of a factorial combination of three methods of input management [high (HIGH), i.e., conventional tillage plus full recommended rates of fertilizer and pesticides; reduced (RED), i.e., conservation tillage plus reduced rates of fertilizer and pesticides; and organic (ORG), i.e., conventional tillage plus N-fixing legumes and non-chemical means of weed and pest control]; and three crop rotation systems with varying levels of cropping diversity [a fallow-based rotation with low crop diversity (LOW); a diversified rotation using annual cereal, oilseed and pulse grains (DAG); and a diversified rotation using annual grains and perennial forages (DAP)]. The study was conducted over the 1996–2007 period on a Dark Brown Chernozemic soil (Typic Boroll) in the Canadian Prairies. As expected, total direct plus indirect energy input was the highest for the HIGH and RED input treatments (3773 MJ ha −1 year −1), and 50% less for ORG management. Most of the energy savings came from the non-use of inorganic fertilizers and pesticides in the ORG management treatments. Further, total energy use was the highest for the DAG treatments (3572 MJ ha −1 year −1), and similar but about 18% lower for the DAP and LOW crop diversity treatments compared to DAG. Thus, overall, the HIGH/DAG and RED/DAG systems had the highest energy requirements (4409 MJ ha −1 year −1) and ORG/DAP had the lowest (1806 MJ ha −1 year −1). Energy output (calorimetric energy content) was typically the highest for the HIGH input treatments (26,541 MJ ha −1 year −1), was about 4% less with RED, and 37% less with ORG management. The latter reflected the lower crop yields obtained with organic management. Similarly, energy output was the highest for the DAP treatments (25,008 MJ ha −1 year −1), about 5% less for DAG, and 20% less for the LOW crop diversity treatments. The higher energy output with the DAP treatments largely reflected that the entire harvested biomass of the forage crops was included in energy output, while for grain crops only the seed was included. The straw and crop residues from annual crops were returned to the land to protect the soil from erosion and to maintain soil organic matter as this is the recommended practice in this semi-arid region. In contrast to energy output and to net energy produced, energy use efficiency (measured as yield of grain plus forage produced per unit of energy input or as energy output/energy input ratio) was the highest for the ORG input treatments (497 kg of harvested production GJ −1 of energy input, and an energy output/energy input ratio of 8.8). We obtained lower, but generally similar energy use efficiency for the HIGH and RED input treatments (392 kg GJ −1 and ratio of 7.1). Thus, overall, ORG/DAP was the most energy efficient cropping system, while RED/LOW and RED/DAG generally ranked the lowest in energy use efficiency. Our findings support the current movement of producers toward ORG management as a means of reducing the reliance on non-renewable energy inputs and improving overall energy use efficiency of their cropping systems. Our results also suggest that moving away from traditional monoculture cereal rotations that employ frequent summer fallowing, toward extended and diversified crop rotations that use reduced tillage methods, although resulting in an increase in energy output, will not significantly reduce the overall reliance on non-renewable energy inputs, nor enhance energy use efficiency, unless perennial legume forages and/or legume grain crops are included in the cropping mix.