We compared efficacy and safety, according to frailty, of elderly patients with relapsed and refractory multiple myeloma (RRMM) treated with lenalidomide and dexamethasone (Rd), for whom bortezomib treatment had failed. Patients, 164 (52.9%) and 146 (47.1%), were classified as non-frail and frail using a simplified frailty scale. The overall response rates (ORR) and survival outcomes were lower in frail than in non-frail patients (ORR: 56.2% vs. 67.7%, P = 0.069; median progression free survival: 13.17 vs. 17.80 months, P = 0.033; median overall survival: 23.00 vs. 36.27 months, P = 0.002, respectively). The number of treatment emergent adverse events in grade 3 or worse was higher in frail than in non-frail patients (41.8% vs. 24.4%, P = 0.002, respectively). In frail patients, independent poor prognostic factors for survival were two or more Charlson comorbidity index (CCI) score, prior to exposure to both bortezomib and thalidomide, and achieved less than partial response In conclusion, frailty could predict clinical outcomes of Rd treatment in elderly patients with RRMM who had failed prior bortezomib. In frail patients, lower CCI in addition to less previous treatment exposure and deep response were associated with better survival.