Previous studies of planet occurrence rates largely relied on photometric stellar characterizations. In this paper, we present planet occurrence rates for mid-type M dwarfs using spectroscopy, parallaxes, and photometry to determine stellar characteristics. Our spectroscopic observations have allowed us to constrain spectral type, temperatures, and in some cases metallicities for 337 out of 561 probable mid-type M dwarfs in the primary Kepler field. We use a random forest classifier to assign a spectral type to the remaining 224 stars. Combining our data with Gaia parallaxes, we compute precise (∼3%) stellar radii and masses, which we use to update planet parameters and planet occurrence rates for Kepler mid-type M dwarfs. Within the Kepler field, there are seven M3 V to M5 V stars which host 13 confirmed planets between 0.5 and 2.5 Earth radii and at orbital periods between 0.5 and 10 days. For this population, we compute a planet occurrence rate of 1.19 +0.70 −0.49 planets per star. For M3 V, M4 V, and M5 V, we compute planet occurrence rates of 0.86 +1.32 −0.68 , 1.36 +2.30 −1.02 , and 3.07 +5.49 −2.49 planets per star, respectively. 2. TARGET SELECTION, OBSERVATIONS, AND DATA REDUCTION Using revised KIC temperature estimates, DC13 identified 202 stars with T eff 3, 300 K. Muirhead et al. (2015) identified 509 probable mid-type M dwarfs from KIC photometry using the following color selection criteria: r − J > 3.2 to identify red objects, and J − K < 0.0555×(r−J)+0.7622 to remove giant stars. The union between these samples produces 561 probable mid-type M dwarfs which we targeted in this study. In total, we observed 337 of the 561 targets (60%), obtaining optical spectra of 327 stars and near-infrared spectra of 82 stars. We have both optical and infrared spectra for 72 targets. Magnitude distributions of our targets in rband and K s -band are shown in Figure 1, including the distributions of targets we observed.