North Polar Cap VLBA Survey (NPCS) proposal. Abstract: We propose to observe the complete sample of ~500 sources in the 0.2 steradians area around the north celestial pole for which simultaneous spectra were measured with RATAN-600 in 2005 and find all compact steep spectrum sources with correlated flux density at X and S bands greater than 20 mJy. Rational: Almost all large VLBA Survey were based on a sample of flat spectrum sources. The sources with spectral index steeper than -0.5, which comprise 80% of source population with flux density > 200 mJy, were usually excluded from the list of observed objects. With completion of VCS1-5 surveys the complete sample of flat spectrum sources at 3/4 of the sky was observed with VLBA. It is tempting to expand conclusions from statistical analysis of this ensemble to the overall population of extragalactic sources. However, we do not know with high confidence how many compact sources among the population of steep spectrum sources with spectral index < 0.5 are missing. We request three 24 hour VLBA observing sessions at X/S band and propose to observe a complete sample of 500 sources with the NVSS flux density at 1.4 GHz greater than 200 mJy in the North Polar Cap region (declination > 75 deg). The purpose of this survey is to find all compact steep spectrum sources in a relatively large field and investigate statistics of this relatively poorly known population. We propose to observe these sources in the mode similar to VLBA Calibrator Survey campaigns, but with longer integration time: 7 minutes. This ensemble of 500 objects was observed with RATAN-600 in 2005, and simultaneous spectra at 2.3, 5.0, 8.0, 11 and 22 GHz were measured for all of these sources during that observing campaign. The objectives of these observations: 1) To get estimates of VLBA correlated flux density at X and S bands if a source has correlated flux > 20 mJy. To get snapshot images, if there is enough detections. 2) To derive coordinates of detected sources at a milliarcsec level of accuracy. 3) To compute statistics of the share of correlated flux density to the total flux density versus spectral index for the steep spectrum population. 4) To compute the distribution of spectral indexes of correlated flux density for the steep spectrum sources population and compare it with the single dish RATAN-600 spectral index. 5) To get the estimate of the total number of compact steep spectrum sources with bright components at a milliarcsecond-scale missed in the VCS1-5 surveys.