MicroRNA (miRNA) genes produce three noncoding RNA products: the long primary transcript (pri-miRNA), the ~70 nucleotide pre-miRNA, and the ~22-nt mature miRNA. Only the mature miRNA is considered to be the functional species of a miRNA gene in recognizing cognate target mRNAs and modulating their expression. However, mature miRNAs are processed from the primary transcript through sequential endonucleolytic steps. As a result, the mature miRNA sequence is present in all three RNA products of a miRNA gene. It has thus been intrinsically difficult to determine the contribution of each miRNA gene product to target repression. In fact, direct functional roles for pri- and pre-miRNAs have never been ruled out. Here we show that pri- and pre-miRNAs may not be mere transitory intermediates of mature miRNA biogenesis. We found that ectopic expression of the C. elegans miRNA gene let-7 (cel-let-7) in human culture cells results in the production of truncated pre- and mature miRNAs that lack the first two 5' nucleotides, one of which is the first nucleotide of the miRNA seed region (nucleotide SD1). We found this nucleotide to be required for repression of target reporters by cel-let-7 in these cells, demonstrating that pri-let-7 may have a direct role in target repression. Further, we show that the nucleotide sequence and structure of both the pri-/pre-let-7 loop and stem regions play a key role in miRNA gene function in reporter assays. In vitro and in vivo analyses indicated the significance of these regions may be in the mediation of a physical interaction between pri-let-7 and target RNAs. These observations suggest that regulatory information encoded in the structured pri-miRNAs, but absent from mature miRNAs, could be directly interpreted for target recognition and repression through RNA:RNA interaction. Intriguingly, some mutations in the loop nucleotide sequence also restored processing of the 5' ends of C. elegans pre- and mature let-7 in culture cells, demonstrating that the pri-/pre-miRNA loop region can also regulate the precision of mature miRNA biogenesis. Importantly, in the presence of functional pre- and mature let-7, cel-let-7 activity in target repression consists of both SD1-independent and SD1-dependent components, implying potential contributions by both pri- and mature let-7. Finally, we interrogated the effects of pri-/pre-let-7 loop mutations on their ability to rescue a let-7 loss-of-function mutant phenotype in C. elegans. Our results indicate decreased significance of these parameters in the control of worm vulval development, although context-dependent differences in mature miRNA biogenesis between heterologous culture and live animals may partially explain this discrepancy. Taken together the work presented here reveals a novel layer of regulatory complexity encoded in long primary miRNAs that may have broad implications in understanding the mechanisms by which miRNA genes control target expression.