Canonical nuclear factor kappaB (NF-κB) activation signals stimulate nuclear translocation of p50:p65, replacing inhibitory p50:p50 with activating complexes on chromatin. C/EBP interaction with p50 homodimers provides an alternative pathway for NF-κB target gene activation, and interaction with p50:p65 may enhance gene activation. We previously found that C/EBPα cooperates with p50, but not p65, to induce Bcl-2 transcription and that C/EBPα induces Nfkb1/p50, but not RelA/p65, transcription. Using p50 and p65 variants containing the FLAG epitope at their N- or C-termini, we now show that C/EBPα, C/EBPα myeloid oncoproteins, or the LAP1, LAP2, or LIP isoforms of C/EBPβ have markedly higher affinity for p50 than for p65. Deletion of the p65 transactivation domain did not increase p65 affinity for C/EBPs, suggesting that unique residues in p50 account for specificity, and clustered mutation of HSDL in the “p50 insert” lacking in p65 weakens interaction. Also, in contrast to Nfkb1 gene deletion, absence of the RelA gene does not reduce Bcl-2 or Cebpa RNA in unstimulated cells or prevent interaction of C/EBPα with the Bcl-2 promoter. Saturating mutagenesis of the C/EBPα basic region identifies R300 and nearby residues, identical in C/EBPβ, as critical for interaction with p50. These findings support the conclusion that C/EBPs activate NF-κB target genes via contact with p50 even in the absence of canonical NF-κB activation and indicate that targeting C/EBP:p50 rather than C/EBP:p65 interaction in the nucleus will prove effective for inflammatory or malignant conditions, alone or synergistically with agents acting in the cytoplasm to reduce canonical NF-κB activation. Mol Cancer Res; 9(10); 1395–405. ©2011 AACR.