Ways in which the ultimate limit of volume titrimetry may be reached are analyzed: the most suitable approach, based essentially upon the use of a bulb burette, allows, in theory, a mechanical precision better than that of the best weight techniques to be obtained, and the various factors influencing this precision are detailed. Several end-point systems potentially sensitive to considerably better than 1 part per million (p.p.m.) are discussed. An attempt at high precision with the iron(II)/dichromate titration, using refined equipment and an amperometric end point, is described: the observed precision was 0.7–1.1 p.p.m. for two experiments of about eight titrations each; the calculated precision, estimated from the resultant of the component precisions, was about 0.9 p.p.m. The ultimate limit of precision of any titrimetric operation, whether by weight or by volume, is shown to be about 0.1 p.p.m.