A systematic investigation of the brazing of Al$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$O$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$ to Kovar$\mathrm{<mrow\; data-mjx-texclass="ORD"><mrow\; data-mjx-texclass="ORD">\circledR </mrow></mrow>}$ (Fe–29Ni–17Co wt.%) using the active braze alloy (ABA) Ag–35.25Cu–1.75Ti wt.% has been undertaken to study the chemical reactions at the interfaces of the joints. The extent to which silica-based secondary phases in the Al$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$O$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$ participate in the reactions at the ABA/Al$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$O$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$ interface has been clarified. Another aspect of this work has been to determine the influence of various brazing parameters, such as the peak temperature, $T$, and time at $T$, $\tau$, on the resultant microstructure. As a consequence, the microstructural evolution of the joints as a function of $T$ and $\tau$ is discussed in some detail. The formation of a Fe$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$Ti layer on the Kovar$\mathrm{<mrow\; data-mjx-texclass="ORD"><mrow\; data-mjx-texclass="ORD">\circledR </mrow></mrow>}$ and its growth, along with adjacent Ni$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$Ti particles in the ABA, dominate the microstructural developments at the ABA/Kovar$\mathrm{<mrow\; data-mjx-texclass="ORD"><mrow\; data-mjx-texclass="ORD">\circledR </mrow></mrow>}$ interface. The presence of Kovar$\mathrm{<mrow\; data-mjx-texclass="ORD"><mrow\; data-mjx-texclass="ORD">\circledR </mrow></mrow>}$ next to the ABA does not change the intrinsic chemical reactions occurring at the ABA/Al$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$O$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$ interface. However, the extent of these reactions is limited if the purity of the Al$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$O$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$ is high, and so it is necessary to have some silica-rich secondary phase in the Al$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$O$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$ to facilitate the formation of a Ti$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$Cu$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$O layer on the Al$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$O$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$. Breakdown of the Ti$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$Cu$\mathrm{<mrow\; data-mjx-texclass="ORD">3</mrow>}$O layer, together with fracture of the Fe$\mathrm{<mrow\; data-mjx-texclass="ORD">2</mrow>}$Ti layer and separation of this layer from the Kovar$\mathrm{<mrow\; data-mjx-texclass="ORD"><mrow\; data-mjx-texclass="ORD">\circledR </mrow></mrow>}$, has been avoided by brazing at temperatures close to the liquidus temperature of the ABA for short periods of time, e.g. for $T$ between 820 and 830 °C and $\tau$ between 2 and 8 min.