We analyzed three years of data from the Kepler space mission to derive rotation periods of main-sequence stars below 6500K. Our automated autocorrelation-based method detected rotation periods between 0.2 and 70 days for 34030 (25.6%) of the 133030 main-sequence Kepler targets (excluding known eclipsing binaries and Kepler Objects of Interest), making this the largest sample of stellar rotation periods to date. In this paper we consider the detailed features of the now well-populated period-temperature distribution and demonstrate that the period bimodality, first seen by McQuillan et al. (2013MNRAS.432.1203M) in the M-dwarf sample, persists to higher masses, becoming less visible above 0.6M_{sun}_. We show that these results are globally consistent with the existing ground-based rotation-period data and find that the upper envelope of the period distribution is broadly consistent with a gyrochronological age of 4.5Gyr, based on the isochrones of Barnes (2007ApJ...669.1167B), Mamajek, & Hillenbrand (Cat. J/ApJ/687/1264) and Meibom et al. (Cat. J/ApJ/695/679). We also performed a detailed comparison of our results to those of Reinhold et al. (Cat. J/A+A/560/A4) and Nielsen et al. (Cat. J/A+A/557/L10), who measured rotation periods of field stars observed by Kepler. We examined the amplitude of periodic variability for the stars with detection rotation periods, and found a typical range between ~950 ppm (5th percentile) and ~22700ppm (95th percentile), with a median of ~5600ppm. We found typically higher amplitudes for shorter periods and lower effective temperatures, with an excess of low-amplitude stars above ~5400K.

Cone search capability for table J/ApJS/211/24/table1 (The rotation period measurements for the 34030 stars presented in this work)

Cone search capability for table J/ApJS/211/24/table2 (Details of the 99000 stars with no significant period detection)