&NA;Eight normal dogs were studied and then made hydrocephalic (kaolin induced) to demonstrate and compare the use of R131ISA and99mTcDTPA in cerebrospinal fluid (CSF) bulk flow studies. The radioisotopes were simultaneously injected into the CSF space in the cisterna magna of normal dogs and into the lateral ventricles of hydrocephalic dogs after an equal volume of CSF had been removed. The average dose of RISA was 100 μCi, and for TcDTPA it was 7 mCi. A bag of lactated Ringer's solution (500 ml) was used as a control phantom after being injected with the same dose of isotope as used in the animals. Brain scans were obtained with the gamma ray camera at 4 and 24 hours after injection. We used four criteria to evaluate results: clinical state of the animal, CSF ventricular pressures with challenges, percentage of retained isotope in counts per minute, and ventricular size at postmortem examination. The results demonstrate that the amount of isotope retention over 20 hours and the clinical state of the dog are better correlated when RISA is used than when TcDTPA is used. The study indicates that RISA is more reliable than TcDTPA as an indicator of CSF bulk flow in both normal and hydrocephalic dogs.