Background: Mortality and disease risk assessments consider body mass index (BMI), among other parameters. Nowadays optimal BMI is discussed controversially as risk assessments are usually performed using BMI of arbitrarily chosen age points. A more comprehensive approach could be based on BMI history. However, longitudinal studies investigating BMI are rare. Objectives: To determine pragmatically different weight history patterns over forty years. Design: Longitudinal study with four follow-ups over forty years, elucidating risk factors for peripheral vessel diseases as the original goal. Subjects: There were 343 male subjects whose weight was measured both at baseline and at follow-ups. Results: Based on pragmatic methods the following eight patterns were found: "stable" (24.8%), "stable and increasing" (28.6%), "stable and decreasing" (9.0%), "hill" (10.0%), "valley" (7.0%), "yo-yo" (14.0%), continually "increasing" (6.4%%), and continually "decreasing" (0.3%). In subjects over 45 years at baseline, stable patterns were most frequent (42%), and descending patterns became more prominent. Conclusions: The determination of different weight history patterns in a longitudinal study is possible with the use of a pragmatic procedure. Applying such weight history patterns to the mortality risk assessment of overweight could add new aspects to that risk assessment.