We present the timing and spectral analyses of the XMM-Newton data on the 17-Myr-old, nearby radio pulsar B0950+08. This observation revealed pulsations of the X-ray flux of PSR B0950+08 at its radio period, P ≃ 253 ms. The pulse shape and pulsed fraction are apparently different at lower and higher energies of the observed 0.2-10 keV energy range, which suggests that the radiation cannot be explained by a single emission mechanism. The X-ray spectrum of the pulsar can be fitted with a power-law model with a photon index Γ = 1.75 ± 0.15 and an (isotropic) luminosity L X = (9.8 ± 0.2) × 10 29 erg s −1 in the 0.2-10 keV. Better fits are obtained with two-component, power-law plus thermal, models with Γ = 1.30 ± 0.10 and L X = (9.7 ± 0.1) × 10 29 erg s −1 for the power-law component that presumably originates from the pulsar's magnetosphere. The thermal component, dominating at E < ∼ 0.7 keV, can be interpreted as radiation from heated polar caps on the neutron star surface covered with a hydrogen atmosphere. The inferred effective temperature, radius, and bolometric luminosity of the polar caps are T pc ≈ 1 MK, R pc ≈ 250 m, and L pc ≈ 3 × 10 29 erg s −1 . Optical through X-ray nonthermal spectrum of the pulsar can be described as a single power-law with Γ = 1.3-1.4 for the two-component X-ray fit. The ratio of the nonthermal Xray (1-10 keV) luminosity to the nonthermal optical (4000-9000Å) luminosity, ≈ 360, is within the range of 10 2 -10 3 observed for younger pulsars, which suggests that the magnetospheric X-ray and optical emissions are powered by the same mechanism in all pulsars, independent of age and spin-down power. Assuming a standard neutron star radius, the upper limit on the temperature of the bulk of the neutron star surface, inferred from the optical and X-ray data, is about 0.15 MK. We also analyze X-ray observations of several other old pulsars, B2224+65, J2043+2740, B0628-28, B1813-36, B1929+10, and B0823+26, and compare their properties with the those of PSR B0950+08.