The population of massive black holes (MBHs) in dwarf galaxies is elusive, but fundamentally important to understand the co-evolution of black holes with their hosts and the formation of the first collapsed objects in the Universe. While some progress was made in determining the X-ray detected fraction of MBHs in dwarfs, with typical values ranging from 0% to 6%, their overall active fraction, A, is still largely unconstrained. Here, we develop a theoretical model to predict the multi-wavelength active fraction of MBHs in dwarf galaxies starting from first principles and based on physical properties of the host, namely its stellar mass and angular momentum content. We find multi-wavelength active fractions for MBHs, accreting at typically low rates, ranging from 5% to 22%, and increasing with the stellar mass of the host as A ∼ (log 10 M ) 4.5 . If dwarfs are characterized by low-metallicity environments, the active fraction may reach ∼ 30% for the most massive hosts. For galaxies with stellar mass in the range 10 7 < M [ M ] < 10 10 , our predictions are in agreement with occupation fractions derived from simulations and semi-analytical models. Additionally, we provide a fitting formula to predict the probability of finding an active MBH in a dwarf galaxy from observationally-derived data. This model will be instrumental to guide future observational efforts to find MBHs in dwarfs. JWST, in particular, will play a crucial role in detecting MBHs in dwarfs, possibly uncovering active fractions ∼ 3 larger than current X-ray surveys.