We examine the effect of overtone interference on fundamental‐mode Love‐wave phase measurements made using single‐station and array‐based techniques at 25–100 s periods. For single‐station teleseismic measurements on USArray Transportable Array data, the contamination effects are small, less than 1% of the path‐averaged phase velocity, consistent with previous studies. Single‐station amplitude measurements provide complementary constraints on the interference pattern. For array‐based measurements on the same data set, contamination effects are much larger: up to ∼10% of the phase velocity for two‐station measurements and up to ∼20% for mini‐array measurements. The interference pattern for single‐station measurements from shallow earthquakes can largely be explained by interactions between only two modes, the fundamental mode and the first higher mode. This interpretation is confirmed using measurements on both mode‐summation synthetic waveforms for a 1D Earth model and synthetic waveforms calculated using SPECFEM3D Globe and a 3D Earth model. Array‐based phase measurements are calculated from differences of the single‐station phase delays, and we demonstrate that the overtone interference pattern for array‐based measurements can be approximated using gradients of the single‐station interference pattern with distance. This relationship can lead to an overall bias to higher phase velocities when combined with common quality selection and data‐reduction procedures for array measurements. Our results indicate that array‐based Love‐wave phase measurements must be carefully scrutinized for overtone contamination and suggest the possibility of new approaches for measuring overtone phase velocities.