We investigate the birth and evolution of Galactic isolated radio pulsars. We begin by estimating their birth space velocity distribution from proper-motion measurements of Brisken and coworkers. We find no evidence for multimodality of the distribution and favor one in which the absolute one-dimensional velocity components are exponentially distributed and with a three-dimensional mean velocity of 380+40–60km/s. We then proceed with a Monte Carlo-based population synthesis, modeling the birth properties of the pulsars, their time evolution, and their detection in the Parkes and Swinburne Multibeam surveys. We present a population model that appears generally consistent with the observations. Our results suggest that pulsars are born in the spiral arms, with a galactocentric radial distribution that is well described by the functional form proposed by Yusifov & Küük, in which the pulsar surface density peaks at radius ∼3 kpc. The birth spin period distribution extends to several hundred milliseconds, with no evidence of multimodality. Models that assume the radio luminosities of pulsars to be independent of the spin periods and period derivatives are inadequate, as they lead to the detection of too many old simulated pulsars in our simulations. Dithered radio luminosities proportional to the square root of the spin-down luminosity accommodate the observations well and provide a natural mechanism for the pulsars to dim uniformly as they approach the death line, avoiding an observed pileup on the latter. There is no evidence for significant torque decay (due to magnetic field decay or otherwise) over the lifetime of the pulsars as radio sources (∼100 Myr). Finally, we estimate the pulsar birthrate and total number of pulsars in the Galaxy.