The occurrence of the zoonotic protozoan parasite Toxoplasma gondii in marine mammals remains a poorly understood phenomenon. In this study, samples from 631 marine mammal species and 34 European otters (Lutra lutra), stranded on the coasts of Scotland, Belgium, the Netherlands and Germany, were tested for the presence of T. gondii. Brain samples were analysed by polymerase chain reaction (PCR) for detection of parasite DNA. Blood and muscle fluid samples were tested for specific antibodies using a modified agglutination test (MAT), a commercial multi-species enzyme-linked immunosorbent assay (ELISA) and an immunofluorescence assay (IFA). Out of 213 animals tested by PCR, only two harbour porpoise (Phocoena phocoena) cerebrum samples, obtained from animals stranded on the Dutch coast, tested positive. The serological results showed a wide variation depending on the test used. Using a cut-off value of 1/40 dilution in MAT, 141 out of 292 animals (41%) were positive. Using IFA, 30 out of 244 tested samples (12%) were positive at a 1/50 dilution. The commercial ELISA yielded 7% positives with a cut-off of the sample-to-positive (S/P) ratio ≥ 50; and 12% when the cut-off was set at S/P ratio ≥ 20. The high number of positives in MAT may be an overestimation due to the high degree of haemolysis of the samples and/or the presence of lipids. The ELISA results could be an underestimation due to the use of a multispecies conjugate. Our results confirm the presence of T. gondii in marine mammals in the Netherlands and show exposure to the parasite in both the North Sea and the Eastern Atlantic Ocean. We also highlight the limitations of the tests used to diagnose T. gondii in stranded marine mammals.