We present a tomographic cosmological weak lensing analysis of the HST COSMOS Survey. Applying our lensing-optimized data reduction, principal component interpolation for the ACS PSF, and improved modelling of charge-transfer inefficiency, we measure
a lensing signal which is consistent with pure gravitational modes and no significant shape systematics. We carefully estimate the statistical uncertainty from simulated COSMOS-like fields obtained from ray-tracing through the Millennium Simulation. We test our pipeline on simulated space-based data, recalibrate non-linear power spectrum corrections using the ray-tracing, employ photometric redshifts to reduce potential contamination by intrinsic galaxy alignments, and marginalize over systematic uncertainties. We find that the lensing signal scales with redshift as expected from General Relativity for a concordance LCDM cosmology, including the full cross-correlations between different redshift bins. For a flat LCDM cosmology, we measure sigma_8(Omega_m/0.3)^0.51=0.75+-0.08 from lensing, in perfect agreement with WMAP-5, yielding joint constraints Omega_m=0.266+0.025-0.023, sigma_8=0.802+0.028-0.029 (all 68% conf.). Dropping the assumption of flatness and using HST Key Project and BBN priors only, we find a negative deceleration parameter q_0 at 94.3% conf. from the tomographic lensing analysis, providing independent evidence for the accelerated expansion of the Universe. For a flat wCDM cosmology and prior w in [-2,0], we obtain w<-0.41 (90% conf.). Our dark energy constraints are still relatively weak solely due to the limited area of COSMOS. However, they provide an important demonstration for the usefulness of tomographic weak lensing measurements from space. (abridged)
