In earlier greenhouse and field studies, we found that upper leaves of Phragmites australis (common reed) and Spartina alterniflora (cord grass) tended to have lower levels of metals than lower leaves. Since lower leaves are produced earlier in the season and are older than upper leaves, there are two possibilities: (1) leaves produced earlier in the season have more metals than those produced later and/or (2) individual leaves continue to accumulate metals during their lifespan. We tested these two hypotheses by numbering all leaves as they emerged from both species of plants growing in Cu-, Pb-, or Zn-contaminated vermiculite in a greenhouse. We removed tissue from the new odd-numbered leaves for metal analysis, and then analyzed the remaining half-leaves later in the season when they had senesced. We could thus compare initial and final metal levels in individual leaves and compare the final metal levels in leaves that emerged early (low numbered leaves) with those that emerged later. We found that hypothesis 2 was supported, since initial levels in individual leaves tended to be lower than the final levels. Hypothesis 1 was not generally supported for either species. We also found very great variation in both initial and final metal levels in leaves on the same plant and in different plants of the same species exposed to the same metal concentration. This suggests that scientists who are sampling leaves of these species in the field should collect many leaves and pool them to get accurate data, since individual leaves are not representative of the plant as a whole.