We studied intra-annual wood density profiles of Douglas-fir tree rings (Pseudotsuga menziesii [Mirb.] Franco) in southwestern Germany. Growth rate varies differently over time throughout the circumference of trees. This leads to differences in wood formation, which can be observed in the shape of the density profiles of the same tree ring measured in different radial directions. Due to this spatial variation in density profiles, we need a reliable method to determine an average profile, which preserves the common characteristics of the data. To this end, we developed a multiple interval-based curve alignment (MICA) procedure. It identifies characteristic points within the profiles such as minima, maxima and inflection points. These reference points are shifted gradually against each other within a proportionally defined base line interval. Using our progressive alignment approach, we are able to calculate an average profile that represents very well the characteristics of all measured curves of a specific tree ring. We applied the procedure to get year-specific average profiles using various trees. This results in representative mean density profiles that preserves the density variations common to all aligned profiles. Individual noise is reduced thereby enabling the analysis of the impact of weather variations on wood density.