ABSTRACT Dwarf galaxies are dark matter (DM) dominated and therefore promising targets for the search for weakly interacting massive particles (WIMPs), which are well-known candidates for DM. The annihilation of WIMPs produces ultrarelativistic cosmic ray electrons and positrons that emit synchrotron radiation in the presence of magnetic fields. For typical magnetic field strengths (few μG) and $\mathcal {O}$(GeV–TeV) WIMP masses, this emission peaks at hundreds of MHz. Here, we use the non-detection of 150-MHz radio continuum emission from the dwarf spheroidal galaxy Canes Venatici I with the Low-Frequency Array to derive constraints on the annihilation cross-section of WIMPs into primary electron–positron and other fundamental particle–antiparticle pairs. Our main underlying assumption is that the transport of the cosmic rays can be described by the diffusion approximation, thus requiring a non-zero magnetic field strength with small-scale structure. In particular, by adopting magnetic field strengths of $\mathcal {O}(1\, μ$G) and diffusion coefficients $∼ \!10^{27}~\rm cm^2\, s^{-1}$, we obtain limits that are comparable with those set by the Fermi Large Area Telescope using gamma-ray observations of this particular galaxy. Assuming s-wave annihilation and WIMPs making up 100 per cent of the DM density, our benchmark limits exclude several thermal WIMP realizations in the [2, 20]-GeV mass range. We caution, however, that our limits for the cross-section are subject to enormous uncertainties that we also quantitatively assess. In particular, variations on the propagation parameters or on the DM halo can shift our limits up by several orders of magnitude (in the pessimistic scenario).