The cornerstones of diagnosis of heartworm (HW) in dogs are the detection of circulating antigen from adult female Dirofilaria immitis or the visualization of microfilariae in whole blood. These tests are less sensitive in cats because of the feline immune response leading to low numbers of adult worms, but heartworm antibody tests are also licensed for use in cats. HW antibodies in cats are detectable when there has at least been larval development in the tissues, but positive antibody tests cannot distinguish between current and previous larval infections; thus, cats with positive antibody test results are considered currently or previously infected with D. immitis. The aim of the present study was to use multiple HW diagnostic modalities to maximize detection of infection in dogs and cats at high risk of infection and to compare infection prevalence between these two hosts. Blood samples collected from 100 stray dogs and 100 stray cats at Florida animal shelters were tested for HW antigen (before and after heat treatment) and microfilariae; cats were also tested for HW antibody. Dogs were significantly (P = 0.0001) more likely to be diagnosed with adult HW infection (28 %; 95 % CI: 20.1–37.6%) when compared with cats (4 %; 95 % CI: 1.6–10.2%) on the basis of positive antigen and microfilariae test results. Cats with current or previous adult, immature adult, or larval HW infections comprised 19 % (95 % CI: 12.4%–27.9%) of the feline population, which was not significantly different (P = 0.1) from the prevalence of adult D. immitis infection in dogs. Testing unprotected cats for heartworm antibodies demonstrated a similar, high risk of infection to the matched unprotected dog population in Florida, which supports the use of HW preventives in cats in areas where HW transmission occurs.