You can use the MySQL client:

$ mysql --user genome --host genome-mysql.cse.ucsc.edu -D mm9 -N -e  "select chrom, txStart, txEnd, name, strand from refGene WHERE name like 'NR_%'" | sort-bed - > ncRNA.mm9.bed

If you're working with mm10:

$ mysql --user genome --host genome-mysql.cse.ucsc.edu -D mm10 -N -e  "select chrom, txStart, txEnd, name, strand from refGene WHERE name like 'NR_%'" | sort-bed - > ncRNA.mm10.bed

Once you have BED files, you can use bedops or bedmap to relate ncRNAs to annotation tables from other sources.