In plants, microRNAs (miRNAs) regulate their mRNA targets by precisely guiding cleavages between the 10th and 11th nucleotides in the complementary regions. High-throughput sequencing-based methods, such as PARE or degradome profiling coupled with a computational analysis of the sequencing data, have recently been developed for identifying miRNA targets on a genome-wide scale. The existing algorithms limit the number of mismatches between a miRNA and its targets and strictly do not allow a mismatch or G:U Wobble pair at the position 10 or 11. However, evidences from recent studies suggest that cleavable targets with more mismatches exist indicating that a relaxed criterion can find additional miRNA targets. In order to identify targets including the ones with weak complementarities from degradome data, we developed a computational method called SeqTar that allows more mismatches and critically mismatch or G:U pair at the position 10 or 11. Precisely, two statistics were introduced in SeqTar, one to measure the alignment between miRNA and its target and the other to quantify the abundance of reads at the center of the miRNA complementary site. By applying SeqTar to publicly available degradome data sets from Arabidopsis and rice, we identified a substantial number of novel targets for conserved and non-conserved miRNAs in addition to the reported ones. Furthermore, using RLM 5'-RACE assay, we experimentally verified 12 of the novel miRNA targets (6 each in Arabidopsis and rice), of which some have more than 4 mismatches and have mismatches or G:U pairs at the position 10 or 11 in the miRNA complementary sites. Thus, SeqTar is an effective method for identifying miRNA targets in plants using degradome data sets.