Fourier domain (FD) techniques have increasingly gained attention in optical coherence tomography (OCT). This is primarily due to their demonstrated sensitivity of two to three orders of magnitude over conventional time-domain techniques. FDOCT images are subject to two primary sources of artifacts. First, a complex conjugate ambiguity arises because the Fourier transform of the real-valued spectral interferometric signal is Hermitian symmetric. This ambiguity leads to artifactual superposition of reflectors at positive and negative pathlength differences between the sample and reference reflectors. Second, noninterferometric and sample autocorrelation terms appear at dc, obscuring reflectors at zero pathlength difference. We show that heterodyne detection in swept-source OCT (SSOCT) enables the resolution of complex conjugate ambiguity and the removal of noninterferometric and autocorrelation artifacts. We also show that complex conjugate ambiguity resolution via frequency shifting circumvents fall-off induced by finite source linewidth in SSOCT when samples are shifted to large pathlength differences. We describe an efficient heterodyne SSOCT design that enables compensation of power losses from frequency-shifting elements. Last, we demonstrate this technique, coupled with wavenumber triggering and electronic demodulation, for in vivo imaging of the human anterior eye segment.