There are only a few reports that investigate the measurement of functional penile length and compare the methods used for this purpose. The aim of this study was to determine the correlation between the measurement techniques and to determine an appropriate method that correlates the erect state most closely. Measurements of penile length in 200 healthy men were made during three different states--flaccid, erect, and stretched--of the penis. The results were evaluated using descriptive statistics, paired t-test, correlation coefficients, and regression equations. The correlation between the flaccid and the erect lengths was 71.2%, and it was 80.2% between the stretched and the erect lengths (p < 0.01). Using regression equations, the degree of accuracy in estimating the erect length by using the stretched length was 65.3% (R2 = determination coefficient; p < 0.01) and was 50.8% by using the flaccid length (p < 0.01). When both the flaccid and the stretched lengths were considered together, the predictability of erect length was R2 = 65.5%. The results obtained using different statistical methods showed that the most accurate results can be obtained by considering stretched penile length, whereas flaccid length had little importance in determining erect penile length. In conclusion, the stretched penile length measurement technique is highly recommended for the accurate prediction of the erect penile length.