Crop residue decomposition not only is mainly driven by, but also affects, soil microorganisms. However, soil microbial responses to legume crops are usually studied only in one subsequent crop. We compared the soil microbial effects of pea (Pisum sativa L.) and faba bean (Vicia faba L.) pulse crops (grown for seed) with faba green manure (GM) and chickling vetch (Lathyrus sativus L.) GM crops in three subsequent crops. Soil microbial biomass C (MBC), β-glucosidase enzyme activity, and bacterial physiological (C substrate utilization) diversity were measured in the summer (rhizosphere and bulk soil) and fall (bulk soil) in all subsequent crops: wheat (Triticum aestivum L.), canola (Brassica napus L.), and barley (Hordeum vulgare L.). Residues of faba bean (grown for GM, herein called faba GM, or for seed, herein called faba bean) usually resulted in the most soil MBC and β-glucosidase activity relative to the other residues. Faba and vetch GM residues increased bulk soil MBC or β-glucosidase enzyme activity more than pulse crop residues in the first and (or) third subsequent crops. Soil MBC and β-glucosidase activities were often positively correlated with initial crop residue N concentrations and negatively correlated with initial C:N ratios or C concentrations. Bacterial physiological diversity was the least responsive to crop residues and was affected differently by sampling time. β-Glucosidase activity was always greater in the fall after crop harvest than in summer. Therefore, β-glucosidase activity was a more sensitive and consistent biological indicator of crop residue effects, and perhaps soil health, than MBC or bacterial physiological diversity.