Direct blood smear examination (using 0.05 ml of whole blood) detected 168 (80.9%) of 204 microfilaremic canine blood samples as determined by the modified Knott test for microfilariae (mff) of Dirofilaria immitis (using 1 ml of whole blood). Direct smear examination detected all of 134 microfilaremias greater than 50 mff ml −1, but only 31 of 70 (44.3%) microfilaremias having less than 50 mff ml −1. In a separate retrospective query of a database of 963 dogs with necropsy-confirmed heartworm infections, 834 (86.6%) were positive by the DiroCHEK heartworm antigen test, and 504 (52.3%) were microfilaremic by the modified Knott test. Only 2 (0.4%) of the microfilaremic dogs were DiroCHEK negative and another 18 (3.6%) were very weak positives. Although these microfilaremic dogs were not tested by direct smear, only one of the two DiroCHEK-negative and six of 18 weakly DiroCHEK-positive dogs had microfilaremias so low that a direct smear may have given a false negative result. Significant adverse reactions to either diethylcarbamazine or the macrolide endectocides have not been reported for microfilaremias less than 500 mff ml −1, thus substitution of the direct smear for a concentration test for mff, such as the modified Knott test or membrane filtration, does not appear to increase the risk of an unexpected adverse reaction to heartworm prophylactic drugs. Such a substitution results in only a very slight decrease (on the order of 0.1%) in the overall sensitivity of heartworm screening, provided a test for mff is run concurrently with an antigen test. If a test for mff is the only screening test used, then substitution of a direct smear for a concentration test may decrease the sensitivity of heartworm screening by nearly 20%, depending on the prevalence of low level microfilaremias in the population of dogs tested.