Abstract: Computational systems may be provided to assist with compression garment recommendations. A measurement of a person's body part may be received, and a compression expected to be applied may be computed. The expected compression may be based upon an arm measurement and upon a known relationship between a size of the garment, such as a stretched circumference, and the compression applied to an object that stretches the garment to that circumference. Graphs and formulas may be used to represent physical properties of a compression garment by assigning compression levels for ranges of relevant circumferences for a multitude of points along a limb. Information about needs of a patient, patient's condition, compression sensitivity, desired compression class, or desired use may be considered as part of a computational analysis to assist in fitting garments. The disclosed recommendation systems may assist clinicians, patients, consumers, or other users in selecting suitable off-the-shelf compression garments.