1. To extend the classification of respiratory neurons based on active membrane properties and discharge patterns to include responses to respiratory modulators, we have studied the effect of thyrotropin-releasing hormone (TRH, 1-5 microM) on the spontaneous respiratory-related neural activity in a thick brain stem slice preparation from the newborn mouse. The action of TRH on the respiratory output from the slice was investigated by recordings from the XII nerve. Cellular responses to TRH were investigated using whole cell recordings from hypoglossal motoneurons and three types of inspiratory neurons located in the rostral ventrolateral part of the slice. 2. Bath-applied TRH (1 microM) decreased the time between inspiratory discharges recorded on the XII nerve from 12.3 +/- 3.3 s to 4.9 +/- 1.1 s (n = 28; means +/- SD), i.e., caused an approximate threefold increase in the respiratory frequency. The coefficient of variation of the time between the inspiratory discharges decreased by one-half. Thus the respiratory output became more stable in response to TRH. The duration of the inspiratory discharges increased from 474 +/- 108 ms to 679 +/- 114 ms, and the amplitude decreased by 24%. An increase in the interdischarge noise on the XII nerve was recorded in the early phase of the TRH application. 3. Anatomically identified hypoglossal motoneurons (7 cells) responded to bath applied TRH with a depolarization eliciting spikes between the inspiratory potentials. The depolarization was accompanied by an increase in spontaneous excitatory synaptic activity that disappeared late during the TRH application. The duration of the inspiratory potentials was increased, indicating that the hypoglossal motoneurons received a longer duration synaptic input from the respiratory rhythm generator. 4. Type-1 inspiratory neurons showed a prolonged depolarization (3 cells), a transient depolarization (2 cells), or no change in membrane potential (2 cells) during 10 min of continued superfusion with a TRH-containing solution. The duration of the inspiratory potentials was increased during the TRH superfusion. With tetrodoxin (TTX, 1 microM) present in the superfusing solution TRH induced a prolonged depolarization (3 cells) or a transient depolarization (1 cell), demonstrating that type-1 inspiratory neurons are depolarized postsynaptically by TRH. The input resistance was not changed during the depolarizing response to TRH. 5. Type-2 inspiratory neurons showed a transient depolarization (7 cells) in response to bath-applied TRH. The duration of the inspiratory potentials was increased markedly during TRH. The transient depolarization was not the result of a postsynaptic action of TRH, because type-2 neurons (9 cells) showed no depolarization to TRH with TTX present in the superfusing solution. 6. Type-3 inspiratory neurons showed a transient depolarization (4 cells) with a partial recovery of the membrane potential late during the TRH application. The duration of the inspiratory potentials increased markedly during TRH. Four cells showed a transient depolarization with an increase in input resistance during TRH with TTX present in the superfusing solution. Thus type-3 neurons are depolarized postsynaptically by TRH. 7. We conclude that TRH increases the frequency of the respiratory rhythm in newborn mice through an action at the level of the brain stem.(ABSTRACT TRUNCATED AT 250 WORDS)