Approximately 40 per cent of older adults living at home are at risk of malnutrition. To detect this condition, body mass index is often used as a tool. The standard way of measuring body height is in a standing position. Height decreases with age and further, many older adults are not physically capable to undergo standing height measures. As height is one of the components in BMI it can be affected by body height decrease which in turn will affect the estimated prevalence of malnutrition risk. There are many different ways to estimate height by proxy measures. The aim of this study was to investigate whether proxy measurements could prove to be useful when estimating height in Swedish community-dwelling older adults. Home visits were made to 51 men and 51 women, who had participated in the randomized, single-blinded health-promoting intervention study, Elderly in the Risk Zone, living in the urban district Örgryte-Härlanda in Gothenburg, Sweden. Body weight, standing height, recumbent height, knee-height and demi-span was measured. An interview was also conducted, retrieving information regarding e.g. height around 20 years of age. T-test and Wilcoxon rank tests were performed and to further examine the results regression analyses and Bland & Altman-plots were conducted. The result showed that between 20 years of age to present age the men had decreased, in body height, on average 3,8 cm and the women 4,9 cm. In both the regression analysis and in the Bland & Altman plots, recumbent height and youth height seem to best conform to standing measured height. In the men, a negative correlation was found between the difference standing and knee-height measure of body height compared to mean values of the two measures in the Bland & Altman plots, though it was the only measurement that did not show any group mean statistical significant difference from standing height by t-test. No negative or positive correlation was seen in the women by the Bland & Altman plots. Demi-span gave an underestimation of body height in both genders. The present results show that body height seems to decrease with age and that besides standing, the best proxy measure is recumbent height otherwise that right knee-height could be used. Different height measurements could affect the BMI classification. Though we need to learn more about what affects the height decrease with age and what proxy measures are reliable. It would be desirable that a larger study would be conducted.