The primary aim of this thesis was to evaluate and improve upon the methods of estimating and measuring small solute clearances in chronic kidney disease. The first part of this thesis addresses the difficult issue of when dialysis should be started in end stage renal disease.  Survival of patients after starting dialysis was assessed retrospectively and no benefit was observed in starting dialysis at a higher level of renal function when patient survival was expressed without the effect of lead-time bias. The relationship between residual renal function, reported uraemic symptoms and the decision to start dialysis for end stage renal failure was assessed in 150 predialysis patients.  In the modern era, when symptoms can be controlled with medication, the decision to start dialysis was based mainly on the level of renal function. The concept of the stop dialysate flow (SDF) method of assessing haemodialysis adequacy was introduced and validated.  This method involves stopping the dialysate flow after the dialysis session and waiting 5 minutes before taking the blood sample.  The work in this thesis demonstrates that the SDF method is more accurate and earlier to perform than other methods of post-dialysis urea sampling. A prediction formula was created and validated that uses the 5-minute blood sample from the SDF method to estimate equilibrated 30-minute urea.  This can then be expressed as equilibrated urea reduction ratio (URR) and Kt/V to increase the utility of the SDF method. The SDF method and prediction formula was shown also to apply in patients receiving haemodiafiltration.