The relationship between osteoarthritis (OA) structural change and pain is complex. Surgical models of OA in rodents are often rapid in onset, limiting mechanistic utility and translational validity. We aimed to investigate the effect of refining surgical small rodent models of OA on both joint pathology and pain behaviour. Adult male C57BL/6 mice (n = 76, 10-11 weeks of age at time of surgery) underwent either traditional (transection of the medial meniscotibial ligament [MMTL]) or modified (MMTL left intact, transection of the coronary ligaments) DMM surgery, or sham surgery. Adult male Sprague Dawley rats (n = 76, weight 175-199g) underwent either modified meniscal transection (MMNX) surgery (transection of the medial meniscus whilst the medial collateral ligament is left intact) or sham surgery. Pain behaviours (weight bearing asymmetry [in mice and rats] and paw withdrawal thresholds [in rats]) were measured pre-surgery and weekly up to 16 weeks post-surgery. Post-mortem knee joints were scored for cartilage damage, synovitis, and osteophyte size. There was a significant increase in weight bearing asymmetry from 13 weeks following traditional, but not modified, DMM surgery when compared to sham operated mice. Both traditional and modified DMM surgery led to similar joint pathology. There was significant pain behaviour from 6 weeks following MMNX model compared to sham operated control rats. Synovitis was significant 4 weeks after MMNX surgery, whereas significant chondropathy was first evident 8 weeks post-surgery, compared to sham controls. Pain behaviour is not always present despite significant changes in medial tibial plateau cartilage damage and synovitis, reflecting the heterogeneity seen in human OA. The development of a slowly progressing surgical model of OA pain in the rat suggests that synovitis precedes pain behaviour and that chondropathy is evident later, providing the foundations for future mechanistic studies into the disease.