OBJECTIVE: To determine proviral load in bovine leukemia virus (BLV)-infected cattle with and without persistent lymphocytosis to assess the potential of transmitting the virus. ANIMALS: Cattle in 6 dairy herds. PROCEDURES: Blood samples from infected cows were evaluated 3 times at 6-month intervals for determination of proviral load via PCR assay, serologic results via ELISA, and hematologic status via differential cell counts. RESULTS: Infected cattle were classified into lymphocytotic and nonlymphocytotic groups. Lymphocytotic cattle consistently had > 100,000 copies of integrated provirus/mug of DNA (ie, high proviral load) in peripheral blood leukocytes. Titers of antibodies against BLVgp51 and BLVp24 indicated a strong immune response. Nonlymphocytotic cattle comprised 2 subgroups: a group with high proviral load and strong immune response, and a group with a weaker immune response, mostly against BLVp24, and a proviral load of < 100 copies/microg of DNA (ie, low proviral load). CONCLUSIONS AND CLINICAL RELEVANCE: Results emphasized the importance of characterizing nonlymphocytotic BLV-infected cattle during eradication programs. The risk of transmitting BLV infection from nonlymphocytotic cattle may differ depending on the proviral load. Nonlymphocytotic cattle with high proviral load could be efficient transmitters (as efficient as lymphocytotic cattle), whereas nonlymphocytotic cattle with low proviral load could be inefficient transmitters under standard husbandry conditions. Because most cattle with low proviral load do not develop anti-BLVp24 antibodies, it appears that lack of an anti-BLVp24 antibody response may be a good marker of this condition.