In this Recommendation, an evaluation methodology is presented for the characterization of low-rate digital speech (or more generally source) coder performance with non-voice signals. The types of non-voice signals considered include voiceband data, network signalling, circuit continuity tones, and dual-tone multi-frequency signalling. Also addressed is the measurement of input/output delay and convergence time. Finally, appropriate test configurations, performance metrics, and test procedures are defined.