Infinite impulse response (IIR) digital filters do not have intrinsic stability; thus, stability constraints should be imposed on their practical designs. The stability-triangle is a necessary-sufficient linear stability condition for IIR digital filters. In order to make good use of the stability-triangle condition, a sequential minimization procedure based on secondorder factor is employed to convert the constrained least-squares design of IIR digital filters into a sequence of constrained least-squares sub-problems, each of which optimizes only one second-order denominator factor while keeping all the rest denominator factors unchanged. Design examples show that the proposed procedure provides better filtering performances than other existing methods.