The ideal serum level of phosphate in patients on dialysis, and the benefits of controlling levels of phosphate in serum remain unclear despite observational studies that associate phosphate levels with mortality. In the absence of robust data from trials, current guidelines are necessarily based on opinion. Oral phosphate binders are required by the majority of patients on dialysis, and all of these binders can control serum levels of phosphate to similar degrees. Patient preference and adherence to prescribed therapy is at least as important as the efficacy of the prescribed binder. Avoidance of calcium-containing binders has become accepted practice where the alternatives are affordable, but incontrovertible evidence in favor of this approach is lacking. Use of sevelamer and lanthanum avoids calcium loading, but at considerable financial cost and with no reliable patient outcome data to prove their value. Additional approaches to aid control of serum levels of phosphate include blockade of gastrointestinal phosphate absorption and possibly binding of salivary phosphate. Importantly, the role of phosphate control in determining patient outcomes must be quantified, which is likely to require a large randomized, controlled study of two levels of phosphate control. Without such a study we will continue to rely on observational data with all its uncertainties and potential to mislead.