Vitamin D status is usually assessed by measuring the serum 25-hydroxyvitamin D (25(OH)D) concentration. This mainly depends on sunshine exposure, nutrition and age. Interlaboratory variation may hamper comparison between results from different populations. This study reports cross-calibration of the 25(OH)D assays of five laboratories. In study 1, serum 25(OH)D was measured with three different assays in 104 serum samples from a large vitamin D supplementation study. The mean serum 25(OH)D level was 80% higher when measured by competitive protein binding (CPB) assay than by high-performance liquid chromatography (HPLC), while radioimmunoassay (RIA) gave intermediate values. The highest correlation was observed between RIA and HPLC (r = 0.84, p < 0.01). Of the serum 25(OH)D values in the lowest quartile by HPLC, 25% were not recognized by CPB and 21% were not recognized by RIA as belonging to the lowest quartile. In study 2, the five laboratories analyzed serum 25(OH)D in eight serum samples covering the concentration range very low to high, with live different assays. The differences between the mean values for serum 25(OH)D between the laboratories with the highest and lowest values was 38%. The ranking order of individual samples according to the serum 25(OH)D value was very similar in all laboratories. The results show that 25(OH)D values from different laboratories can not be assumed to be comparable unless a careful cross-calibration has been performed.