The priming principle was investigated in the rat phrenic nerve-diaphragm preparation stimulated continuously at 0.2 Hz. Tubocurarine was added to the organ bath as either a single (non-primed) or a divided (primed) dose. Priming consisted of 15% or 20% of the final dose, with priming intervals of 5 or 10 min. Priming decreased significantly the time to 80% block and was associated with mild neuromuscular block. A simple model adequately predicted the time to 50% and 80% block, using the same diffusion constant for both primed and non-primed conditions. The onset of neuromuscular block, with and without priming, depended mostly upon the distribution of the drug to its site(s) of action.