Reforestation is effective in restoring ecosystem functions and enhancing ecosystem services of degraded land. The three most commonly employed reforestation methods of natural reforestation, artificial reforestation with native Masson pine, and introduced slash pine plantations were equally successful in biomass yield in southern China. However, it is not known whether soil ecosystem functions, such as nitrogen (N) cycling, are also successfully restored. Here, we employed a functional microarray to illustrate soil N-cycling. The composition of N-cycling genes in soils varied significantly with reforestation method and varied with constructive species identity and plant diversity. Artificial reforestation had less superior organization of N-cycling genes than natural reforestation, as indicated by the less diverse and less stable pathways to perform the biogeochemical N-cycling processes. Besides, artificial reforestation had lower functional potential (abundance of ammonification, denitrification, assimilatory, and dissimilatory nitrate reduction to ammonium genes) in soils than natural method. Evaluations of the abundance and interactions of N-cycling genes in soils showed that plantations, especially artificial reforestation with slash pine plantations, possessed a smaller range of ecosystem functions that provide a less diverse array of N-related substrates and nutrients to microbial communities compared with natural restoration. This might lead to a lower independence of N-cycling, which indicated a higher risk of N release in plantations. The unfavorable N-cycling conditions in plantations were corroborated by the lower contents of available N, ammonium N, and nitrate N. These findings demonstrate that reforestation methods could have broad regional and possibly global implications for N-cycling.