Cosmological constraints on the sum of neutrino masses and on the effective number of neutrino species in standard and nonstandard scenarios are computed using the most recent available cosmological data. Our cosmological data sets include the measurement of the baryonic acoustic oscillation (BAO) feature in the data release 9 CMASS sample of the baryon oscillation spectroscopic survey. We study in detail the different degeneracies among the parameters, as well as the impact of the different data sets used in the analyses. When considering bounds on the sum of the three active neutrino masses, the information in the BAO signal from galaxy clustering measurements is approximately equally powerful as the shape information from the matter power spectrum. The most stringent bound we find is Sigma m(nu) < 0.32 eV at 95% C.L. When nonstandard neutrino scenarios with N-eff massless or massive neutrino species are examined, power spectrum shape measurements provide slightly better bounds than the BAO signal only, due to the breaking of parameter degeneracies. Cosmic microwave background data from high multipoles from the South Pole Telescope turns out to be crucial for extracting the number of effective neutrino species. Recent baryon oscillation spectroscopic survey data combined with cosmic microwave background and Hubble Space Telescope measurements give N-eff = 3.66(-0.21-0.69)(+0.20+0.73) in the massless neutrino scenario, and similar results are obtained in the massive case. The evidence for extra radiation N-eff > 3 often claimed in the literature therefore remains at the 2 sigma level when considering up-to-date cosmological data sets. Measurements from the Wilkinson Microwave Anisotropy Probe combined with a prior on the Hubble parameter from the Hubble Space Telescope are very powerful in constraining either the sum of the three active neutrino masses or the number of massless neutrino species. If the former two parameters are allowed to freely vary, however, the bounds from the combination of these two cosmological probes get worse by an order of magnitude.