Nitrogen (N) excretion and emissions can be reduced in fattening pigs by reducing dietary crude protein (CP) level. Effects of this strategy are well documented for moderate CP reduction, but little literature exists on further CP reduction made possible by free isoleucine, histidine, and leucine. This trial evaluated the effects of two levels of reduction in CP on growth performance, N balance, and gaseous emissions. Forty-eight pigs were allocated to 12 gaseous emission-measuring chambers at 28kg live weight. Three dietary treatments (CTRL; -1.2pt; -2.4pt), with a CP content, respectively, of 18.1%, 16.9%, and 15.0% in phase 1 (28-48kg) and 16.1%, 15.0%, and 13.8% in phase 2 (48-80kg), were fed ad libitum. Growth performance was recorded for each phase. Body lipid and protein composition were analyzed by dual-energy X-ray absorptiometry in two animals per chamber at the beginning and end of the trial. These results were used to calculate the N balance. Slurry volume and composition were measured at the end of the trial. Ammonia, methane, and nitrous oxide emissions were recorded continuously. Data were analyzed with a general linear model including the linear and quadratic effect of CP reduction and phase as fixed effects. A trend for a quadratic effect of CP reduction on feed intake was observed (P = 0.085) with a decrease from CTRL to -1.2pt and then an increase with the -2.4pt treatment. Daily gain and gain to feed were improved by the low-CP diets in phase 1 but were degraded in phase 2 (phase × CP interaction, P < 0.001) while body composition was not affected. Nitrogen excretion decreased with CP reduction (linear effect, P = 0.023) but tended to plateau with the low-CP diets (quadratic effect, P = 0.081). Methane emissions per kg of gain were reduced with CP reduction (linear effect, P = 0.031). Ammonia emissions decreased with CP reduction, mostly in phase 1 (phase × CP interaction, P = 0.015); however, the emission factor (g N-NH3 / g urinary N) was not affected. Nitrous oxide emissions were not affected by CP reduction and its emission factor (g N-N2O / g N) tended to increase (linear effect, P = 0.07). Slurry ammonia-N decreased (linear effect, P < 0.023; quadratic effect, P = 0.049), but other slurry components were not significantly impacted. In this trial, a plateau in N excretion and emission reduction was reached with the -2.4pt treatment. Nevertheless, it is important to study such a reduction in conditions closer to commercial ones.