Total respiratory resistance (Rrs) was measured in six anesthetized dogs with two different methods: the single-breath (SB) method, which provides the time constant of the system during a relaxed expiration and the forced oscillation (FO) method, which uses a pseudorandom noise signal applied at the airway opening. The comparison was made in three conditions: before muscle paralysis (A), after muscle paralysis (B), and after tracheal banding (C). In conditions A and B the two computed resistances correlated very well with each other (r = 0.98). No systematic difference between Rrs values obtained with the two methods was found. In condition C the respiratory resistance was clearly nonlinear from the flow-volume curves during SB and could be described with Rohrer's equation: Rrs = K1 X V + K2 X V2, where K1 and K2 are Kohrer's constant and V is flow. Rrs measured with FO was not frequency dependent during tracheal banding (C) and was virtually equivalent to K1. Since the FO method uses low flows as the input of the respiratory system and K1 could be ascribed to laminar flow, the numerical matching appears reasonable and tends to reinforce the validity of both methods of measurement. We conclude that, for the normal respiratory system, FO and SB methods are approximately equivalent. In the presence of a markedly alinear central airway resistance with normal lungs, the SB method appears to provide a more adequate description of the flow-resistive properties of the system.