Many cladistic systematists consider hypotheses of ancestor-descendant relationship untestable in fossils. Although fossil preservation is never complete enough to know which fossil population is actually ancestral, theoretically it is possible to sample the ancestral population. If all potentially ancestral populations were sampled, then some members of the actual ancestral population must be contained in the sample. Preservation of most vertebrate and macroinvertebrate fossils is notoriously spotty. However, marine microfossil sequences are the most complete available in the fossil record. Since most planktonic microorganisms have biogeographic distributions that closely correspond to the extent of major water masses, it is possible to sample each population by sampling the water masses. Lack of preservation due to non-deposition, erosion, or dissolution can generally be recognized. We can tell if absence of a form from an area is real or due to lack of preservation. Piston cores from the entire world ocean are available for all the water masses since the late Miocene. Their stratigraphy can be carefully worked out by at least three independent means of correlation, and in some cases these cores faithfully record events spaced only hundreds of years apart. Thus, in certain pelagic microfossils it is possible to sample all populations that have existed through millions of years, and be confident that no forms of interest remain unsampled. Therefore, the marine microfossil record can meet the criteria necessary to recognize ancestors. An example from the Radiolaria is discussed.