Nuclear relaxation is a sensitive monitor of rotational dynamics in spin-label EPR. It also contributes competing saturation transfer pathways in T1-exchange spectroscopy, and the determination of paramagnetic relaxation enhancement in site-directed spin labelling. A survey shows that the definition of nitrogen nuclear relaxation rate Wn commonly used in the CW-EPR literature for 14N-nitroxyl spin labels is inconsistent with that currently adopted in time-resolved EPR measurements of saturation recovery. Redefinition of the normalised 14N spin-lattice relaxation rate, b = Wn/(2We), preserves the expressions used for CW-EPR, whilst rendering them consistent with expressions for saturation recovery rates in pulsed EPR. Furthermore, values routinely quoted for nuclear relaxation times that are deduced from EPR spectral diffusion rates in 14N-nitroxyl spin labels do not accord with conventional analysis of spin-lattice relaxation in this three-level system. Expressions for CW-saturation EPR with the revised definitions are summarised. Data on nitrogen nuclear spin-lattice relaxation times are compiled according to the three-level scheme for 14N-relaxation: T1 n = 1/Wn. Results are compared and contrasted with those for the two-level 15N-nitroxide system.