This study investigated the nitrogen (N) acquisition from soil and insect capture during the growth of three species of pitcher plants, Nepenthes mirabilis, Cephalotus follicularis and Darlingtonia californica. N-15/N-14 natural abundance ratios (delta(15)N) of plants and pitchers of different age, non-carnivorous reference plants, and insect prey were used to estimate proportional contributions of insects to the N contest of leaves and whole plants. Young Nepethes leaves (phyllodes) carrying closed pitchers comprised major sinks for N and developed mainly from insect N captured elsewhere on the plant. Their delta(15)N values of up to 7.2 parts per thousand were higher than the average delta(15)N value of captured insects delta(15)N value 5.3 parts per thousand), In leaves carrying old that are acting as a N source, the delta(15)N decreased to 3.0 parts per thousand indicating either an increasing contribution of soil N to those plant parts which in fact captured the insects or N gain from N-2 fixation by microorganisms which may exist in old pitchers, The delta(15)N value of N in water collected from old pitchers was 1.2 parts per thousand and contained free amino acids. The fraction of insect N in young and old pitchers and their associated leaves decreased from 1.0 to 0.3 mg g(-1) This fraction decreased further with the size of the investigated tiller. Nepenthes contained on average 61.5 +/- 7.6% (mean +/- SD, range 50-71%) insect N Eased on the N content of a whole tiller, In the absence of suitable non-carnivorous reference plants for Cephalotus, delta(15)N values were assessed across a developmental sequence from young plants lacking pitchers to large adults with lip to 38 pitchers. The data indicated dependence on soil N until 4 pitchers had opened. Beyond that stage, plant size increased with the number of catching pitchers but the fraction of soil N remained high. Large Cephalotus plants were estimated to derive 26 +/- 5.9% (mean +/- SD of the three largest plants; range: 19-30%) of the N from insects. In Cephalotus we observed an increased delta(15)N value in sink versus source pitchers of about 1.2 parts per thousand on average. Source and sink pitchers of Darlingtonia had a similar delta(15)N value, but plant N in this species showed delta(15)N signals closer to that of insect N than in either Cephalotus or Nepenthes. Insect N contributed 76.4 +/- 8.4% (range 57-90%) to total pitcher N content. The data suggest complex patterns of partitioning of insect and soil-derived N between source and sink regions in pitcher plants and possibly higher dependence on insect N than recorded elsewhere for Drosera species.