Electric-dipole matrix elements for 6p −nd, n = 5, 6, 7 transitions in cesium are calculated using a relativistic all-order method. The resulting matrix elements are used to evaluate 5d lifetimes and 6p polarizabilities. The data are compared with experimental lifetime and polarizability measurements made by different groups. Domination of the 6p scalar polarizabilities by 5d − 6p dipole matrix elements facilitates an exacting consistency check of 5d lifetime and 6p polarizability data. Values of 5d − 6p matrix elements obtained from experimental 5d lifetime data are found to be inconsistent with those inferred from 6p polarizabilities derived from experimental Stark shift data. Our ab initio calculated 6p polarizabilities agree well with experimental determinations.PACS numbers: 31.15. Ar, 32.70.Cs, 32.10.Dk, 31.15.Dv The understanding of the accuracy of ab initio calculations in cesium is vital for the analysis of the Cs parity nonconservation (PNC) experiment [1]. In 1999, motivated by a number of recent high-precision experiments, Bennett and Wieman [2] reanalyzed the agreement of theoretical calculations and experimental data for a number of Cs atomic properties and reduced the previous theoretical uncertainty in the PNC amplitude by a factor of two. Utilizing measurements of the tensor transition polarizability, β, reported in same work, they demonstrated a 2.5σ discrepancy between the value of the weak charge Q W predicted by the Standard Model and that derived from the Cs PNC experiment. Although several papers (for example, [3,4,5,6,7,8,9]), have addressed this disagreement since 1999, the issue of the accuracy of ab initio calculations in Cs continues to be of interest.In this work, we investigate the radiative properties of Cs 6p − nd transitions. Although these do not bear directly on PNC experiments done to date, they have been the subject of careful experimental investigation, and thus provide benchmarks for precise comparison of theory and experiment. In particular, there exist two independent measurements of the lifetimes of the 5d states [10,11], which do not agree within their stated uncertainties. There also exist several experimental determinations of the 6p − 6s Stark shifts which allow to infer the values of polarizabilities of the 6p states [12,13,14]. Here we show that ab initio theory can check the mutual consistency of 5d lifetime and 6p polarizability data, with an accuracy of about 1%. We find the lifetime and polarizability results to be inconsistent at this level. Our calculations agree with the experimental values of 6p polarizabilities, but deviate from both determinations of the 5d lifetimes. We suggest that further experiments are desirable in order to clarify this issue. In addition, * Electronic address: msafrono@physics.udel.edu; Current address: Department of Physics and Astronomy, University of Delaware, Newark, Delaware, 19716 understanding of the accuracy of the 5d state properties in Cs is germane to the ongoing PNC experiment in isoelectronic Ba + [15], since the ...