Thyroid cancer is increasing in incidence worldwide. While outcomes are generally good, up to 25% of patients suffer recurrence, and this has a significant impact on their quality of life and life expectancy. We hypothesised that thyroid tumours which recur display a distinct pattern of driver events, present on initial histology. Controlled-access TCGA data on thyroid cancer were downloaded and whole exome sequencing data analysed. An analysis pipeline utilising Platypus, Annovar and SIFT/PolyPhen2/MutationTaster filtering was performed in n=43 recurrent patients. This identified mutations in genes including Inosine-5(-monophosphate dehydrogenase 2 (IMPDH2), 6-Phosphofructo-2-Kinase/Fructose-2,6-Biphosphatase 4 (PFKFB4) and Dicer 1 ribonuclease type III (DICER1). In silico analysis suggested these variants to be pathogenic and therefore they were recapitulated using site-directed mutagenesis. Cellular migration, invasion and subcellular localisation were investigated in cell lines representing the most common background driver mutations of papillary thyroid cancer (TPC1 cells (RET/PTC mutation); SW1736 (BRAFV600E); Cal62 (Ras)). IMPDH2 mutation significantly increased cell migration at 4, 8 and 24hrs vs. WT (P=0.0068, P=0.0008, P=0.0088, respectively), and DICER1 mutation induced increased cell migration at 24 h vs. vector-only (P=0.0094) in TPC1 cells. An analysis of RNA and microRNA expression levels was performed, comparing recurrent (n=43) to non-recurrent (n=457) TCGA patients. In the RNA analysis genes involved in matrix adhesion and thyroid cancer pathogenesis were most differentially expressed in recurrent patients, including fibronectin 1 (FN1) and α3 integrin (ITGA3). Overexpression of FN1 increased cell migration (TPC1s P=0.007; SW1736s P=0.0001; Cal62 P=0.01) and knockdown of ITGA3 decreased cell migration at 24 h (TPC1 P=0.001 SW1736 P=0.0003, Cal62 P=0.0056) but not cell proliferation. MicroRNA analysis highlighted miR 221, 486 and 1179 as miRs significantly differentially expressed in recurrence. We propose that altered RNA and miRNA expression levels may be key to predicting thyroid cancer recurrence.