This research measures the conditions of spinal scoliosis of college students and analyses the factors that affect spinal scoliosis. From September 15, 2008 to October 30, 2008, the study was conducted with 142 C College students and measured the degrees of spinal scoliosis. The analysis was evaluated according to the Pneumex Analysis program. The collected data was compiled into statistics using the SPSS PC+ 12.0 program. The results were as follows: 1. It appeared that in the cases of C curve, reverse C curve, S curve, and the normal, there was a statistical significance(p=.010) between the conditions of spinal scoliosis of the two genders. 2. It appeared that the comparison among the three groups according to the characteristics of the subjects showed a statistical significance in spinal length(p=.000) in the C curve group, age(p=.019) and spinal length(p=.004) in the reverse C curve group, and spinal length (p=.035), difference in apparent leg length(p=.019), and difference in true leg length(p=.012) in the S curve group. 3. It appeared that in the normal group without spinal scoliosis, the results of comparing the differences in apparent true leg length and spinal length according to the characteristics of the subjects showed a statistical significance in spinal length and gender(p=.000), age (p=.008), height(p=.000), and whether or not there was a regular exercise over once per week(p=.001). 4. On the correlations between C curve, reverse C curve, and S curve and differences in apparent true leg length and spinal length, it appeared that: (a) As the difference in apparent leg length gets bigger the difference in true leg length gets big in C curve(r=.551, p=.000). (b) As the difference in apparent leg length gets bigger the difference in true leg length gets big in reverse C curve(r=.511, p=.006). (c) There is no correlation in S curve.