Physical function is an important outcome in chronic kidney disease (CKD). We aimed to establish the reliability, validity, and the "minimal detectable change" (MDC) of several common tests used in renal rehabilitation and research. In a repeated measures design, 41 patients with CKD not requiring dialysis (stage 3b to 5) were assessed at an interval of 6 weeks. The tests were the incremental shuttle walk test (ISWT), "sit-to-stand" (STS) test, estimated 1 repetition maximum for quadriceps strength (e1RM), and VO by cardiopulmonary exercise testing (CPET). Reliability was assessed using intraclass correlation coefficient and Bland-Altman analysis, and absolute reliability by standard error of measurement and MDC. The ISWT, STS-60, e1RM, and CPET had "good" to "excellent" reliability (0.973, 0.927, 0.927, and 0.866), respectively. STS-5 reliability was poor (0.676). The MDC is ISWT, 20 m; STS-5, 7.5 s; STS-60, 4 reps; e1RM, 6.4 kg; VO, 2.8 ml/kg/min. There was strong correlation between the ISWT and VO (r = 0.73 and 0.74). While there was poor correlation between the STS-5 and e1RM (r = 0.14 and 0.47), better correlation was seen between STS-5 and ISWT (r = 0.55 and 0.74). In conclusion, the ISWT, STS-60, e1RM, and CPET are reliable tests of function in CKD. The ISWT is a valid means of exercise capacity. The MDC can help researchers and rehabilitation professionals interpret changes following an intervention.