We present a final unified catalogue for the 7C survey at 151 MHz with resolution 70x70cosec(dec) arcsec^2^. This has been constructed by amalgamating the existing catalogues derived from individual fields imaged at this resolution and eliminating redundancy in regions of mutual overlap. This is a non-trivial procedure because the flux in multiple-component sources may be fitted differently on alternative images, owing, for example, to differences in local noise and beam distortion. The final catalogue thus produced contains 43683 sources. Separate final catalogues have been published for the 7C Galactic Plane survey (7CG, see Cat. J/MNRAS/294/607) and the lower-resolution survey of the low-declination strip 9h<RA<16h, 20deg<Dec<35deg (Cat. J/MNRAS/282/779). The individual catalogues for about 40 of the regions contributing to the total have already been published, together with full details of the methodology, in MNRAS or A&AS: Lacy et al. 1995, MNRAS, 276, 614 (=1995MNRAS.276..614L) (#92 below) Visser et al. 1995, A&AS, 110, 419 (=1995A&AS..110..419V) (#93 below) Pooley et al. 1998, MNRAS, 298, 637 (=1998MNRAS.298..637P) (#94-96 below) Riley et al. 1999, MNRAS, 306, 31 (=1999MNRAS.306...31R) (# 1-33 below) and these data are also available via the MRAO website at http://www.mrao.cam.ac.uk/surveys/7C/ Individual catalogues for the remaining 58 regions by Riley et al. (#34-91 below) were released electronically via the MRAO website in November 2001. These include a re-analysis of data originally published in rather a different parametrization by McGilchrist et al. 1990, MNRAS, 246, 110 (=1990MNRAS.246..110M) The regions re-analyzed are those numbered #41,44,59,60,62 and 63 below and they supersede McGilchrist's 1990 publication. The RAxDec coverage, average rms noise, flux density of the faintest source listed and completeness limit for each of the individual regions contributing to the final catalogue are given in the table "regions.dat". 1-sigma errors on the listed source positions may be approximated by: RA..error(arcsec) = SQRT(1.0**2 + (32/SNR)**2) Dec.error(arcsec) = Kcosec(dec) x (RA error) where (approx) K= 1.0 around dec 70, increasing to 1.3 below dec 50, and 1-sigma errors on the listed flux densities may be approximated by: Error on S beam-fitted(Jy) = SQRT(0.03**2 + SNR**-2) x S(Jy) Error on SINT(Jy) = 1.5 x SQRT(0.03**2 + SNR**-2) x SINT(Jy) where SNR, S and SINT correspond to the columns denoted by those names in the byte-by-byte description below. For multi-component sources the SNR for the brightest component is used to estimate the SINT error. For further details of the surveys and data analysis procedures please refer to the published papers referenced at the head of this file and references therein.