It is already clear that alternative splicing has an extremely important role in expanding the protein diversity. Comparative study of the codon usage patterns of alternatively and commonly spliced genes may thereby be necessary. In this paper, the patterns of codon usage bias of two kinds of human genes, alternatively spliced genes and commonly spliced genes, were formulated through analyzing 344 non-redundant protein coding sequences from alternatively spliced genes (188183 codons) and 386 from commonly spliced genes (223116 codons) in human chromosome (1. Overall) codon usage data analysis indicated that the alternatively spliced genes showed a stronger codon usage bias than commonly spliced genes. Very extensive heterogeneity of G C content in silent third codon position (GC3s) was evident among these genes, and GC3s content of alternatively spliced genes was higher than that of commonly spliced genes. G- or C-ending codons were more abundant in alternatively spliced genes than commonly spliced genes in human chromosome 1. The causation of differences created could be explained by pre-mRNA structural characteristics of alternatively spliced genes influencing their codon usage bias.