We investigate possible resonance effects in the primordial power spectrum using the latest CMB data. These effects are predicted by a wide variety of models and come in two flavors, one where the oscillations are log spaced and one where the oscillations are linearly spaced. We treat the oscillations as perturbations on top of the scale invariant power spectrum. This allows us to significantly improve the search for resonance because it allows us to precompute the transfer functions. We show that the largest error from this simplification comes from the variance in the measurement to the distance of last scattering. In case of log spaced oscillations this introduces a shift in the phase (as opposed to the true phase) while for linear spaced oscillations it changed the effective frequency. We test our code on simulated Planck data, where we are able to recover fiducial input oscillations with O(10−2) level perturbations at the 95% C.L. We use both fiducial as well as nil-simulations to address our confidence. Running a total of 3000 Markov Chain Monte Carlo for each model, we put stringent constraints on the presence of resonance in the WMAP9 and Planck data. We apply existing evidence criteria to qualify our findings.