Using an in situ substratum enrichment method, we assessed the effects of nutrient release from sand substrata on the community structure of attached diatoms in a sand-bottomed, northern Michigan stream. Sand from the stream bottom was washed, sterilized, and then consolidated into small plastic petri dishes with agar solutions enriched with various concentrations of NaNO3 and KH2PO4. Laboratory estimates of nutrient release rates were performed over a 144-h period with uncolonized substrata having different agar and nutrient concentrations. Release rates decreased 10-fold in an exponential fashion with no significant differences between replicates. Agar concentrations had no significant effect on release rates, although the rates were proportional to nutrient concentrations in the substratum. Racks of substrata were installed parallel to the stream current flow and retrived after a 6-wk colonization period. Using multiple comparisons of treatment means for both nutrients and algal taxa, we found no differences between control and NO3 enrichments; significant differences were found, however, between PO4 and NO3 + PO4 (N;P = 25:1) treatment means relative to each other and the control. Diatom biovolume was two times as great on PO4 treatments and four times greater on NO3 + PO4 treatments. Effects of PO4 and NO3 enrichment appear to be taxa specific, with PO4 and NO3 + PO4 treatments favoring Navicula and Nitzschia spp. Control and NO3 treatments were dominated by Cocconeis placentula Ehr. and Achnanthes minutissima Kutz.