A set of controlled high-Reynolds-number experiments has been conducted at the William B. Morgan Large Cavitation Channel (LCC) in Memphis, Tennessee to investigate the friction drag reduction achieved by injecting aqueous poly(ethylene oxide) (PEO) solutions at three different mean molecular weights into the near-zero-pressure-gradient turbulent boundary layer that forms on a smooth flat test surface having a length of nearly 11 m. The test model spanned the 3.05 m width of the LCC test section and had an overall length of 12.9 m. Skin-friction drag was measured with six floating-plate force balances at downstream-distance-based Reynolds numbers as high as 220 million and free stream speeds up to 20 m s-1. For a given polymer type, the level of drag reduction was measured for a range of free stream speeds, polymer injection rates and concentrations of the injected solution. Polymer concentration fields in the near-wall region (0 < y+ < ∼103) were examined at three locations downstream of the injector using near-wall planar laser-induced-fluorescence imaging. The development and extent of drag reduction and polymer mixing are compared to previously reported results using the traditional K-factor scaling. Unlike smaller scale and lower speed experiments, speed dependence is observed in the K-scaled results for the higher molecular weight polymers and it is postulated that this dependence is caused by molecular aggregation and/or flow-induced polymer degradation (chain scission). The evolution of near-wall polymer concentration is divided into three regimes: (i) the development region near the injector where drag reduction increases with downstream distance and the polymer is highly inhomogeneous forming filaments near the wall, (ii) the transitional mixing region where drag reduction starts to decrease as the polymer mixes across the boundary layer and where filaments are less pronounced and (iii) the final region where the polymer mixing and dilution is set by the rate of boundary layer growth. Unlike pipe-flow friction-drag reduction, the asymptotic maximum drag reduction (MDR) either was not reached or did not persist in these experiments. Instead, the nearest approach to MDR was transitory and occurred between the development and transitional regions. The length of the development region was observed to increase monotonically with increasing polymer molecular weight, injection rate, concentration and decreasing free stream speed. And finally, the near-wall polymer concentration is correlated to the measured drag reduction for the three polymer molecular weights in the form of a proposed empirical drag-reduction curve.