The study aimed to (1) assess whether higher vasti (VASTI), gluteus medius (GMED), gluteus maximus (GMAX) and gluteus minimus (GMIN) forces are associated with participant characteristics (lower age, male gender) and clinical characteristics (lower radiographic disease severity, lower symptom severity and higher walking speed); and (2) determine whether hip and knee muscle forces are lower in people with patellofemoral joint (PFJ) osteoarthritis (OA) compared to those without PFJ OA.Sixty participants with PFJ OA and 18 (asymptomatic, no radiographic OA) controls ≥40 years were recruited from the community or via referrals. A three-dimensional musculoskeletal model was used in conjunction with optimisation theory to calculate lower-limb muscle forces during walking. Associations of peak muscle forces with participant and clinical characteristics were conducted using Pearson's r or independent t-tests and between-group comparisons of mean peak muscle forces performed with walking speed as a covariate.Peak muscle forces were not significantly associated with participant, symptomatic or radiographic-specific characteristics. Faster walking speed was associated with higher VASTI muscle force in the PFJ OA (r = 0.495; P < 0.001) and control groups (r = 0.727; P = 0.001) and higher GMAX muscle force (r = 0.593; P = 0.009) in the control group only. Individuals with PFJ OA (N = 60) walked with lower GMED and GMIN muscle forces than controls (N = 18): GMED, mean difference 0.15 [95% confidence interval (CI): 0.01 to 0.29] body weight (BW); GMIN, 0.03 [0.01 to 0.06] BW. No between-group differences were observed in VASTI or GMAX muscle force: VASTI, 0.10 [-0.11 to 0.31] BW; GMAX, 0.01 [-0.11 to 0.09] BW.Individuals with PFJ OA ambulate with lower peak hip abductor muscle forces than their healthy counterparts.