The effects of exercise therapy (ET) on renal function in chronic kidney disease (CKD) remain unclear. In a randomized controlled trial (UMIN-CTR number: UMIN000038415), we investigated whether ET affects renal function in CKD; eligible patients had undergone renal biopsy in the past 3months. We stratified patients by disease (immunoglobulin A [IgA] nephropathy, n = 16; diabetic nephropathy, n = 4; benign nephrosclerosis, n = 13; and other CKD types, n = 13) and randomized them to 12weeks' observation and 24weeks' ET comprising home-based aerobic exercise 3×/week and resistance training 2×/week (intervention group) or usual care (non-intervention group). Primary endpoint was creatinine-based estimated glomerular filtration rate (eGFR) or serum cystatin C-based eGFR (eGFRcys). Secondary endpoints included urinary protein and exercise tolerance. Seventy patients were enrolled, 50 fulfilled the inclusion criteria, but 4 discontinued before randomization. No items significantly differed between week 0 to 24 in either group (intervention group, n = 23; non-intervention group, n = 23) or between groups at week 24 (intention-to-treat population) in the total study population. The eGFRcys slope showed no significant intergroup difference in the observation period, but eGFRcys improved significantly in IgA nephropathy patients (n = 16) in the intervention group (stratified comparison; week 0, 48.3 ± 18.2; week 24, 51.6 ± 17.6; p = 0.043). In these patients, urinary protein was significantly worse at week 24 in the non-intervention group (p = 0.046) and worsened significantly less in the intervention group (p = 0.039). ET did not improve renal function overall in CKD patients but might help maintain renal function in patients with IgA nephropathy.