We report comprehensive 77 Se NMR measurements on a single crystalline sample of the newly discovered FeSe-based high temperature superconductor KxFe2−ySe2 (Tc = 33 K) in a broad temperature range up to 290 K. Despite deviations from the stoichiometric KFe2Se2 composition, we observed 77 Se NMR lineshapes as narrow as 4.5 kHz under a magnetic field applied along the crystal c-axis, and found no evidence for co-existence of magnetic order with superconductivity. On the other hand, the 77 Se NMR lineshape splits into two peaks with equal intensities at all temperatures when we apply the magnetic field along the ab-plane. This suggests that K vacancies may have a superstructure, and that the local symmetry of the Se sites is lower than the tetragonal four-fold symmetry of the average structure. This effect might be a prerequisite for stabilizing the s± symmetry of superconductivity in the absence of the hole bands at the Brillouin zone center. From the increase of NMR linewidth below Tc induced by the Abrikosov lattice of superconducting vortices, we estimate the in-plane penetration depth λ ab ∼ 290 nm, and the carrier concentration ne ∼ 1 × 10 +21 cm −3 . Our Knight shift 77 K data indicate that the uniform spin susceptibility decreases progressively with temperature, in analogy with the case of FeSe (Tc ∼ 9 K) as well as other FeAs high Tc systems. The strong suppression of 77 K observed immediately below Tc for all crystal orientations is consistent with a singlet pairing of Cooper pairs. We don't, however, observe the Hebel-Slichter coherence peak of the nuclear spin-lattice relaxation rate 1/T1 immediately below Tc, expected for conventional BCS s-wave superconductors. In contrast with the case of FeSe, we do not observe evidence for an enhancement of low frequency antiferromagnetic spin fluctuations (AFSF) near Tc in 1/T1T . Instead, 1/T1T exhibits qualitatively the same behavior as overdoped non-superconducting Ba(Fe1−xCox)2As2 with x ∼ 0.14 or greater, where hole bands are missing in the Brillouin zone center. We will discuss the implications of our results on the hitherto unknown mechanism of high temperature superconductivity in FeSe and FeAs systems.