Our aim in this thesis was twofold. The first aim was to describe the potential determinants of muscle strength in patients with knee OA. The second aim was to optimize strength gains in patients with knee osteoarthritis. We performed a scoping review to examine which factors were associated with muscle strength in patients with knee OA. Factors were categorized into three subgroups: basic muscle properties, general and OA-specific factors. Regarding the basic muscle properties , we found convincing evidence for the association between low muscle quality and low muscle strength in patients with knee OA, but not for muscle quantity and muscle activation. With regard to general factors in patients with knee OA, we concluded that older age, being female, higher bodyweight and physical inactivity were found to be associated with lower muscle strength. With respect to the OA-specific factors in patients with knee OA, more severe joint degeneration and higher levels of pain were found to be associated with lower muscle strength. In patients with hip OA, we found a lack of studies across all factors potentially associated with muscle strength. We performed a study cross-sectional study in the OAI progression cohort (n=1390) that showed that dietary protein intake of in 65.1% of the patients with knee OA was below the recommended daily allowance (RDA) of 0.80 g/kg bodyweight for healthy adults. We also showed that lower dietary protein intake was associated with lower muscle strength in patients with knee OA. We performed a study on longitudinal data in the OAI progression cohort (n=1390). We found a decline in muscle strength over a 4-year period in patients with knee OA. Predictors for the observed decline in muscle strength were older age, being female, higher BMI, lower dietary protein intake and radiographic severity. We assessed whether in patients with knee OA (i) high-intensity resistance training leads to an increase in muscle strength, compared to low-intensity resistance training. We found that in patients with knee OA, high-intensity resistance training did not lead to greater improvement in isokinetic muscle strength compared to low-intensity resistance training. Our findings from isokinetic strength data suggest that the dose-response relationship of resistance training (that is, higher training intensities leads to greater improvements than lower training intensities) reported in older adults, does not seem to hold for patients with knee OA. Two other studies confirm these findings with no difference between the high- and low-intensity resistance training groups. Our study also showed that high-intensity resistance training was well tolerated in patients with knee OA. In addition, we showed that resistance training with vitamin D supplementation did not lead to an increase in muscle strength compared to placebo, but this finding was based on underpowered data. We studied isokinetic peak strength and the estimated 1-RM in relation to daily activities. Contrary to our hypothesis, isokinetic peak strength showed stronger associations compared to the estimated 1-RM with all daily activities, including walking, chair rising and stair climbing. These findings support the use of isokinetic peak strength compared to the estimated 1-RM in research in patients with knee OA, particularly in relation to daily activities. Activities requiring muscular endurance, such as walking, were not found to be more strongly associated with 1-RM than with isokinetic muscle strength, contrary to our hypothesis. This may be explained by the fact that the 10-RM test is a multiple repetition to failure-test with only 10 repetitions, which may be not enough to test muscle endurance. Finally, the main results of this thesis were summarized and discussed and directions for further research were provided.