This study examined the relationship between skinfold compressibility and skinfold thickness while proposing a 3-caliper technique for the estimation of uncompressed skinfold thickness. Skinfold and ultrasound measurements of skin-plus-adipose tissue were made at seven sites on 15 females (mean age 22.0 yr) and 13 males (mean age 23.2 yr.) The estimated uncompressed thickness (ZT) determined with the 3-caliper technique correlated well with ultrasound-determined thickness (UT), with the 3-caliper technique offering a valid, simple and inexpensive method of estimating uncompressed skinfold thickness. Using linear regression analyses, skinfold compressibility varied with skinfold thickness. Using a compression percentage [C(Z/10)], both UT and ZT exhibited positive slopes with tissue compressibility. The concept of absolute thickness change (TD) from uncompressed (ZT) to compressed (H10 ) thickness values was adopted for use in covariance analysis in the assessment of gender and site differences in skinfold compressibility and was compared to results of traditional compression percentage analysis of variance. In a second sample of 40 males (mean age 25.7 yr) and 30 females (mean age 25.0 yr), a significant site difference was found using compression percentage ANOVA which was not present using the TD ANCOVA. This difference in results may be due to the fact that compression percentage is linearly related to thickness even though it is included as the denominator in the calculation of compression percentage. It was concluded that the TD covariance method should be used in preference to the compression percentage ANOVA when evaluating skinfold compressibility differences in gender, age and site of measurement. © 1993 Wiley-Liss, Inc.