Abstract: We revisit the issue of superconductivity at the quantum-critical point between a 2D paramagnet and a spin-density-wave (SDW) metal with ordering momentum (pi,pi). This problem is highly non-trivial because the system at criticality displays a non-Fermi liquid behavior and because the effective coupling constant &amp;#955; for the pairing is generally of order one, even when the actual interaction is smaller than fermionic bandwidth. Previous study [M. A. Metlitski, S. Sachdev, Phys.Rev.B 82, 075128 (2010)]has found that the leading renormalization of the pairing vertex contains log2, like in color superconductivity. We analyze the full gap equation and argue that summing up log2 term does not lead to a pairing instability. Yet, superconductivity has no threshold, even if &amp;#955; is set to be small: the subleading log terms give rise to BCS-like Tc&amp;#8733;e&amp;#8722;1/&amp;#955;. We argue that the analogy with BCS is not accidental as superconductivity at a QCP is a Fermi liquid phenomenon -- it comes from fermions which retain Fermi liquid behavior at criticality. We compute Tc for the actual &amp;#955; and find consistency with the numerical results.<br>